Skip to main content
This guide provides deeper explanations of advanced concepts for users who want to understand more about how the Credential Builder works and how to optimize their results. If you’re looking for step-by-step instructions, start with the Building Your Credential guide instead.
This guide is for experienced users who want to master the Credential Builder’s advanced features. It assumes you’ve already built at least one credential and are familiar with the basic workflow.

​
Understanding Optimization Weights

The Credential Builder uses three optimization weights that work together as an interconnected system to balance competing goals when designing credentials. Rather than adjusting individual weights, most users should choose a configuration profile that matches their credential goals.
Critical Understanding: The three weights (Overlap Penalty, Undercoverage Penalty, Section Count Penalty) are not independent dials—they work as a coordinated system. Adjusting one weight without understanding its relationship to the others can lead to poor results. Instead, start by choosing a configuration profile that matches your credential’s purpose.

​
Quick Overview of the Three Weights

Before diving into configuration profiles, here’s a brief explanation of what each weight controls: Overlap Penalty (0-100%): Controls how much the system avoids selecting multiple course sections that teach the same skills. Low values (0-5%) allow skill reinforcement; high values (40%+) aggressively eliminate redundancy. Undercoverage Penalty (0-100%): Controls how aggressively the system tries to meet all job skill requirements. High values (> 60%) prioritize complete coverage; low values (30-40%) accept gaps for other benefits. Section Count Penalty (0-100%): Controls preference for shorter vs. longer credentials. High values (50-70%) favor compact programs; low values (10-30%) allow more sections for comprehensive coverage.
💡 Key Insight: For most workforce-focused credentials, the Overlap Penalty should be LOW (0-5%) and the Undercoverage Penalty should be HIGH (> 60%). This ensures students learn everything required for job readiness, with strategic skill reinforcement across courses.

​
Configuration Profiles: Choose Your Approach

Instead of adjusting individual weights, choose one of these four proven configuration profiles based on your credential’s primary goal:

​
Profile 1: Maximum Job Readiness (Meet ALL Requirements)

THE PRIMARY PROFILE - Use This for Most Workforce Credentials

When to use: Workforce training, professional certifications, employer partnerships, career-focused certificates, industry credentialsGoal: Ensure students learn 100% of required skills to complete proficiency levelsWeight Configuration:
  • Overlap Penalty: 0-5% (Low - allows skill reinforcement)
  • Undercoverage Penalty: > 60% (High - no or few gaps allowed)
  • Section Count Penalty: 10-20% (Low - comprehensive coverage prioritized)
What You Get:
  • Students prepared to meet ALL job requirements
  • Strategic skill reinforcement across multiple courses
  • May result in longer programs (acceptable for job readiness goal)
  • High optimization scores reflecting strong job market alignment
Real-World Example:
A community college creates an RN credential preparing students for the NCLEX-RN exam and employment as registered nurses.Configuration: Overlap 3%, Undercoverage 95%, Section Count 15%Result: System selects 12 course sections ensuring every required nursing skill (patient assessment, medication administration, pharmacology, emergency response, clinical judgment) reaches required proficiency levels. Some skills like “Patient Communication” and “Clinical Documentation” appear in multiple clinical courses—this is intentional and valuable.Outcome: 100% of graduates meet NCLEX requirements. The credential takes 2 years to complete, but comprehensive preparation is non-negotiable for healthcare careers.
When This Profile Shows Low Scores: If you’re using Profile 1 and getting optimization scores below 60%, you don’t have a weights problem—you have a curriculum gap problem. Your course catalog lacks the skills needed to prepare students for your target jobs. The solution is NOT to adjust weights; it’s to develop new courses or choose different target jobs that match your existing curriculum.

​
Profile 2: Balanced Coverage & Efficiency

Standard Certificates & Stackable Credentials

When to use: Standard certificate programs, stackable credentials, general industry alignment, professional developmentGoal: Strong job readiness while keeping programs reasonably compactWeight Configuration:
  • Overlap Penalty: 10-20% (Low-Medium - minimizes redundancy while allowing some reinforcement)
  • Undercoverage Penalty: 50-60% (Medium-High - covers most requirements)
  • Section Count Penalty: 30-40% (Medium - balanced program length)
What You Get:
  • Good job market alignment (70-85% coverage typical)
  • Reasonable program length (students can complete in 1-2 semesters)
  • Minimal skill redundancy with strategic overlap where beneficial
  • Balanced trade-offs between competing goals
Real-World Example:
A professional development program creates a digital marketing certificate for working professionals seeking career advancement.Configuration: Overlap 15%, Undercoverage 65%, Section Count 35%Result: System selects 6-7 course sections covering core marketing skills (social media marketing, content strategy, SEO/SEM, analytics, campaign management). Some foundational skills like “Data Analysis” appear in 2 courses—beneficial reinforcement. Program totals 18-21 credits.Outcome: Students gain strong digital marketing competency in one academic year. May have minor gaps in cutting-edge specialized areas (marketing automation, advanced attribution modeling), but covers 80%+ of entry-level job requirements. Graduates are employable with room to grow on the job.

​
Profile 3: Focused Micro-Credential (ONLY After Coverage Confirmed)

⚠️ ADVANCED OPTIMIZATION - Not a Starting Point

When to use: ONLY after confirming 95%+ coverage with Profile 1, now need to reduce redundancy and program lengthCritical Requirement: Never use Profile 3 as your starting configuration. First confirm high coverage with Profile 1, then switch to Profile 3 to trim excess.Weight Configuration:
  • Overlap Penalty: 25-40% (Medium-High - removes redundancy)
  • Undercoverage Penalty: 50-60% (Medium - maintains most coverage)
  • Section Count Penalty: 50-60% (Medium-High - favors compact programs)
What You Get:
  • Compact, efficient credential (2-4 courses typical)
  • Minimal skill redundancy
  • Reduced program length and cost
  • May sacrifice some depth for brevity
Real-World Example:
A continuing education program creates a focused badge on data visualization after confirming their curriculum provides strong coverage.Phase 1 - Coverage Verification (Profile 1):
  • Configuration: Overlap 5%, Undercoverage 90%, Section Count 15%
  • Result: 6 courses selected, 98% coverage of required visualization skills
Phase 2 - Optimization for Brevity (Profile 3):
  • Configuration: Overlap 35%, Undercoverage 55%, Section Count 60%
  • Result: 3 courses selected, 92% coverage maintained (minor gaps in advanced techniques)
Outcome: Compact 9-credit badge completable in one semester. Students learn core visualization principles (Tableau, PowerBI, design theory, storytelling with data) without excessive redundancy. The slight coverage drop (98% → 92%) is acceptable for a focused badge.
NEVER Use Profile 3 as Starting Point:If you start credential design with Profile 3 settings and get low optimization scores, switching to Profile 1 won’t necessarily fix the problem—the algorithm has already selected sub-optimal sections.Correct Workflow:
  1. Start with Profile 1 → Verify high coverage
  2. If coverage is good (90%+), switch to Profile 3 to trim excess
  3. If coverage is poor, you have curriculum gaps—develop new courses or adjust target jobs

​
Profile 4: Exploratory/Foundational (Partial Coverage OK)

Introductory Credentials & Career Exploration

When to use: Introductory badges, career exploration programs, foundational credentials that explicitly don’t aim for full job readiness, first step in multi-credential pathwayGoal: Introduce students to a field without attempting comprehensive job preparationWeight Configuration:
  • Overlap Penalty: 5-10% (Low - allows reinforcement of foundational skills)
  • Undercoverage Penalty: 30-40% (Low-Medium - explicitly accepts gaps)
  • Section Count Penalty: 60-70% (High - keeps programs very short)
What You Get:
  • Brief introduction to field (2-3 courses typical)
  • Foundational skills only
  • Explicit acknowledgment that further credentials required for job readiness
  • Low barrier to entry encourages exploration
Real-World Example:
A community college creates an introductory cybersecurity badge for students exploring whether to pursue a cybersecurity career.Configuration: Overlap 8%, Undercoverage 35%, Section Count 65%Result: System selects 2-3 course sections covering cybersecurity fundamentals (network basics, threat awareness, security principles). Explicitly does NOT cover advanced requirements like penetration testing, incident response protocols, or cryptographic implementation. Coverage: ~40% of entry-level Security Analyst job requirements.Outcome: Students complete the badge in 6-9 credits (one semester), gaining enough knowledge to decide if cybersecurity interests them. Badge description explicitly states: “This is an introductory credential. For job readiness, continue to our Cybersecurity Technician Certificate.” Acts as first step in credential pathway.
Profile 4 vs. Profile 1: Profile 4 is NOT a “smaller version” of Profile 1. It serves a fundamentally different purpose:
  • Profile 1: Job readiness (students can work in the field)
  • Profile 4: Career exploration (students learn if they want to work in the field)
Use Profile 4 only when you explicitly want an incomplete, foundational credential that leads to further study.

​
How to Choose the Right Profile

Use this decision framework to select your configuration profile:
1

Ask: Must Students Be 100% Job-Ready?

YES → Profile 1 (Maximum Job Readiness)NO → Continue to Step 2
2

Ask: Is This an Introductory/Exploratory Credential?

YES (students will need additional credentials for job readiness) → Profile 4 (Exploratory/Foundational)NO (this should prepare students for employment) → Continue to Step 3
3

Ask: Do You Already Have 95%+ Coverage?

YES (verified with Profile 1 run) → Profile 3 (Focused Micro-Credential)NO (haven’t verified coverage yet) → Continue to Step 4
4

Default Choice

Profile 2 (Balanced Coverage & Efficiency)Use this for standard certificates balancing job readiness with reasonable program length.
Question 1: Must students be 100% job-ready? Answer: Yes—can’t work as RN without complete preparationProfile Selection: Profile 1 (Maximum Job Readiness)Rationale: Healthcare professions require comprehensive preparation. Gaps in knowledge can harm patients. Use Profile 1 even if it results in longer programs.
Question 1: Must students be 100% job-ready? Answer: No—this is a 2-course introductionQuestion 2: Is this introductory/exploratory? Answer: Yes—helps students explore web developmentProfile Selection: Profile 4 (Exploratory/Foundational)Rationale: Explicitly accepting partial coverage for brief introduction. Students continue to full certificate for job readiness.
Question 1: Must students be 100% job-ready? Answer: Mostly—want strong preparation but some on-the-job learning is normalQuestion 2: Is this introductory/exploratory? Answer: No—should prepare for employmentQuestion 3: Do you have 95%+ coverage? Answer: Haven’t verified yetProfile Selection: Profile 2 (Balanced Coverage & Efficiency)Rationale: Standard workforce certificate balancing thorough preparation with reasonable program length. Can be completed in 1-2 semesters while covering 75-85% of job requirements.

​
The Critical Relationship: Overlap vs. Undercoverage Ratio

Understanding the relationship between Overlap Penalty and Undercoverage Penalty is crucial for credential quality.

​
The Golden Ratio

Undercoverage Penalty should be 3-10x HIGHER than Overlap Penalty This ratio ensures the system prioritizes meeting job requirements over avoiding redundancy. Good Ratio Examples:

Ratio 16:1 (Ideal for Job Readiness)

  • Overlap: 5%
  • Undercoverage: 80%
  • Ratio: 16:1
Result: System teaches all required skills, allows strategic reinforcement. Redundancy is avoided only when it doesn’t compromise coverage.

Ratio 3:1 (Acceptable for Balanced Credentials)

  • Overlap: 20%
  • Undercoverage: 60%
  • Ratio: 3:1
Result: System balances coverage with efficiency. Some skill gaps acceptable to avoid excessive overlap.
Bad Ratio Examples:

Ratio 1:1 (Conflicting Priorities)

  • Overlap: 40%
  • Undercoverage: 40%
  • Ratio: 1:1
Problem: System can’t decide whether to avoid overlap or ensure coverage. Results are unpredictable and often poor.

Ratio 1:2 (Inverted - TERRIBLE)

  • Overlap: 60%
  • Undercoverage: 30%
  • Ratio: 1:2 (inverted!)
Problem: System prioritizes avoiding skill overlap MORE than meeting job requirements. Will skip teaching required skills just to prevent duplication. Catastrophic for job readiness.

​
Why High Overlap Penalty Hurts Job Readiness

When Overlap Penalty is too high relative to Undercoverage Penalty, the system makes counterproductive decisions: Scenario: Two target jobs both require “SQL Database Management” at Level 4 proficiency.
  • Section A teaches SQL to Level 2
  • Section B teaches SQL to Level 3
  • Combined, they would reach Level 4 (meeting requirement)
With Balanced Ratio (Overlap 5%, Undercoverage 80%): System selects BOTH sections because meeting the Level 4 requirement (high priority) is worth accepting some overlap (low penalty). Result: Students learn SQL to required proficiency ✅ With Inverted Ratio (Overlap 70%, Undercoverage 40%): System selects ONLY ONE section (maybe Section B) because the overlap penalty (high priority) outweighs the gap in coverage (low penalty). Result: Students only reach Level 3 SQL—insufficient for job requirements ❌
The Overlap Trap: Setting Overlap Penalty above 40% for job-readiness credentials causes the system to avoid teaching required skills just to prevent duplication. This defeats the entire purpose of the Credential Builder. For 90% of credentials, Overlap Penalty should stay between 0-20%.

​
When Configuration Profiles Don’t Work

If you’ve chosen the appropriate profile but still get poor results, the problem isn’t your weight settings—it’s your data.

​
Symptom: Low Optimization Scores with Profile 1

What it means: Your course catalog cannot teach the skills required for your target jobs. What it’s NOT: A problem with weight configuration. Solutions:
1

Review Skills Coverage Report

Check the embedded report in the Summary tab. Identify which specific skills have 0% coverage or coverage significantly below required levels.
2

Verify Target Job Selection

Are your target jobs realistic given your course catalog? For example, if you’re a community college without medical courses, targeting “Cardiovascular Surgeon” is unrealistic.
3

Identify Curriculum Gaps

Make a list of missing skills. Determine whether:
  • Courses exist but weren’t selected (try manually activating them)
  • Courses exist but syllabi haven’t been processed for skills
  • Courses genuinely don’t exist in your catalog
4

Develop New Curriculum OR Adjust Target Jobs

Either:
  • Develop new courses teaching the missing skills
  • Choose different target jobs that match your existing curriculum
Don’t try to “fix” this by lowering Undercoverage Penalty—that just hides the problem.

​
Symptom: Profiles Produce Nearly Identical Results

What it means: Limited course selection forces the algorithm to choose the same sections regardless of weights. Causes:
  • Very few courses teach required skills
  • Course filters are too restrictive
  • A few “dominant” sections teach many critical skills
Solutions:
1

Remove Course Filters

Try running optimization with no filters. If results dramatically change, your filters were too restrictive.
2

Expand Course Catalog

Add more sections with varied content to give the algorithm more options.
3

Accept the Result

If the same sections consistently appear because they objectively teach critical skills best, accept that this is the optimal solution.

​
Best Practices for Profile-Based Configuration

Don’t: Open Settings tab and immediately start adjusting individual weight sliders.Do: Choose the profile that matches your credential goal, run optimization, then make small refinements if needed.Why: Profiles represent proven, balanced configurations tested across hundreds of credentials. Custom weight combinations often create conflicting priorities.
Before creating any focused or compact credential, run Profile 1 to verify your catalog CAN meet job requirements.Workflow:
  1. Select target jobs
  2. Run optimization with Profile 1 settings
  3. Review coverage in Summary tab
  4. If coverage is 90%+, proceed with your desired profile
  5. If coverage is below 70%, you have curriculum gaps—address before proceeding
Why: Prevents wasting time trying to optimize credentials that can’t succeed due to missing curriculum.
Keep notes on which profile you selected for each credential and why.Example:
Cybersecurity Technician Certificate
Profile: Profile 1 (Maximum Job Readiness)
Rationale: Partnership with local employers requires graduates
           meet 100% of Security Technician job requirements
Weights: Overlap 5%, Undercoverage 90%, Section Count 15%
Optimization Score: 87%
Notes: Initially tried Profile 2 but employers needed complete
       coverage. Accepted longer 12-course program for comprehensive
       preparation.
Why: Institutional memory. Helps you and colleagues understand decisions months or years later.
When designing a new type of credential you haven’t built before, run optimizations with 2-3 different profiles.Example Workflow:
  1. Run Profile 1 → Note score and section count
  2. Run Profile 2 → Note score and section count
  3. Run Profile 3 → Note score and section count
  4. Compare results side-by-side
  5. Choose profile that best serves your goals
Why: Empirical comparison reveals trade-offs and helps you make informed decisions.
Profiles are starting points, not rigid requirements. Small adjustments within a profile are fine.Example:
  • Profile 1 suggests: Overlap 0-5%, Undercoverage >60%, Section Count 10-20%
  • You use: Overlap 5%, Undercoverage 85%, Section Count 18%
Why: Your specific institutional context may benefit from slight tuning. Just maintain the overall balance and ratios.Don’t: Make large departures that change the profile’s character (e.g., setting Overlap to 50% while claiming to use Profile 1).

​
Troubleshooting Profile Results

​
Issue: Profile 1 Shows Only 65% Optimization Score

Diagnosis: Curriculum gaps—your courses can’t teach required skills. NOT the Problem: Weight configuration. Solution:
  1. Review Skills Coverage report (Summary tab)
  2. Identify missing or undercovered skills
  3. Check if courses exist but weren’t selected (try manual activation)
  4. If skills genuinely missing, develop new courses or adjust target jobs
  5. Don’t lower Undercoverage Penalty—that hides the problem without fixing it

​
Issue: Profile 3 Produces Poor Scores

Diagnosis: You’re using Profile 3 as a starting point instead of an optimization step. Solution:
  1. Reset optimization
  2. Run with Profile 1 settings first
  3. Verify you achieve 90%+ coverage with Profile 1
  4. THEN switch to Profile 3 to trim excess
  5. If you can’t achieve high coverage with Profile 1, don’t use Profile 3

​
Issue: All Profiles Produce Similar Results

Diagnosis: Limited course options force same selections regardless of weights. Possible Causes:
  • Course filters too restrictive
  • Catalog lacks variety in certain skill areas
  • A few “dominant” sections teach many critical skills
Solution:
  1. Remove all course filters and run again
  2. If results differ dramatically, filters were too restrictive
  3. If results stay similar, accept that algorithm is finding objectively optimal sections
  4. Consider expanding course catalog with more varied content

​
Issue: Want Job Readiness but Program Too Long

Diagnosis: Conflict between coverage goals and length constraints. Reality Check: You can’t have complete job readiness in 2-3 courses if jobs require 50+ distinct skills. Solutions:
  1. Accept longer program: Use Profile 1, accept the required section count
  2. Split into pathway: Create foundational badge (Profile 4) + advanced certificate (Profile 1)
  3. Reduce target jobs: Narrow focus to fewer, more specific jobs requiring fewer skills
  4. Adjust target jobs: Choose entry-level jobs with fewer skill requirements
Don’t: Set high overlap penalty (40%+) trying to force brevity—this creates coverage gaps and defeats job readiness goals.

​
How the Optimization Algorithm Works (Non-Technical)

The Credential Builder uses a genetic algorithm—a type of artificial intelligence inspired by biological evolution. Here’s how it works in plain language.

​
The Evolution Analogy

Imagine natural selection, but instead of organisms evolving, course section combinations are evolving toward better job market alignment. 1. Population (Initial Generation) The algorithm starts by creating 150 random combinations of course sections (this is the “population size”). Each combination is like a different species trying to survive. Example Population Members:
  • Combination A: Sections 12, 45, 67, 89
  • Combination B: Sections 3, 23, 56, 78, 90
  • Combination C: Sections 5, 15, 25, 35, 45, 55
  • …and 147 more random combinations
2. Fitness Evaluation (Natural Selection) Each combination gets a “fitness score” based on how well it achieves your goals:
  • Skill Coverage: Does it teach the required job skills to required proficiency levels?
  • Overlap: How much redundancy exists between selected sections?
  • Section Count: How many sections does it require?
The fitness function uses your three optimization weights to calculate a single overall score for each combination. 3. Selection (Survival of the Fittest) The algorithm keeps the top-performing combinations and discards the rest. In nature, organisms with better traits survive to reproduce. Here, combinations with better fitness scores “survive” to the next generation. Typically, the top 30-40% of combinations survive each generation. 4. Crossover (Breeding) The surviving combinations are “bred” together to create new combinations. This is called crossover. Example Crossover:
  • Parent A: Sections [12, 45, 67, 89]
  • Parent B: Sections [3, 23, 56, 78, 90]
  • Child 1: Sections [12, 45, 56, 78, 90] (took some from each parent)
  • Child 2: Sections [3, 23, 67, 89] (different mix from the same parents)
This process combines the strengths of good solutions to potentially find even better solutions. 5. Mutation (Random Changes) Occasionally (15% of the time by default), the algorithm makes random changes to combinations. This prevents the algorithm from getting stuck in “local optima”—solutions that seem good but aren’t the best possible. Example Mutation:
  • Original: Sections [12, 45, 67, 89]
  • After mutation: Sections [12, 45, 67, 89, 102] (added a random section)
  • Or: Sections [12, 19, 67, 89] (replaced section 45 with random section 19)
6. Repeat (Evolution Continues) Steps 2-5 repeat for many generations (typically 50-200). Each generation should have better average fitness than the previous one. The best combination from all generations is tracked and ultimately returned as your optimized result.

​
Why Optimization Takes 30-300 Seconds

The algorithm must evaluate thousands of possible combinations:
  • Population size: 150 combinations per generation
  • Generations: 50-200 generations typically
  • Total evaluations: 150 × 100 = 15,000 combinations evaluated
For each combination, the system must:
  1. Calculate skills coverage across all selected sections
  2. Compare coverage to requirements for every target job
  3. Identify overlapping skills between sections
  4. Calculate fitness score using your weight configuration
Factors Affecting Optimization Time:
  • More target jobs = Longer (more skill requirements to evaluate)
  • More available course sections = Longer (larger search space)
  • More complex job skill requirements = Longer (more calculations per combination)
  • Higher generation count = Longer (more iterations)
Typical optimization times:
  • Small credential (1-2 jobs, 20-50 sections available): 30-60 seconds
  • Medium credential (3-5 jobs, 50-200 sections available): 60-180 seconds
  • Large credential (6-10 jobs, 200+ sections available): 180-300 seconds
If optimization takes longer than 5 minutes, consider reducing the number of target jobs or using course filters to limit available sections.

​
Understanding Real-Time Progress

While the optimization runs, you see live updates showing the algorithm’s progress: Generation Counter (e.g., “Generation 45/100”) Shows which iteration the algorithm is on. Progress through generations is your primary completion indicator.
  • Early generations (1-20): Algorithm is exploring diverse solutions, scores may vary widely
  • Middle generations (21-70): Algorithm is converging on good solutions, steady improvement
  • Late generations (71-100): Fine-tuning the best solutions, smaller incremental improvements
Best Fitness Score Shows the fitness score of the best combination found so far. Higher scores are better. This number should generally increase over time, though it may plateau in later generations once the algorithm finds optimal or near-optimal solutions.
If best fitness stops improving after 30-40 generations and stays flat for 20+ more generations, the algorithm has likely found the optimal solution. The remaining generations are confirming there are no better options.
Elapsed Time Shows how long the optimization has been running. Use this to estimate remaining time based on progress percentage. Progress Percentage Visual indicator showing completion status. Based on current generation divided by total generations.

​
What “Generations” Actually Mean

Each generation represents a complete cycle of:
  1. Evaluate fitness of all combinations
  2. Select best performers
  3. Create new combinations through crossover and mutation
  4. Replace poor performers with new combinations
More generations mean more thorough exploration of possible solutions, but also longer runtime. The default (50-200 generations depending on problem complexity) balances thoroughness with reasonable execution time.

​
Why Not Just Manually Design Credentials?

You might wonder: “Why use an algorithm when humans could design credentials?” The Combinatorial Explosion Problem: Even a small credential design problem has millions of possible combinations:
  • 50 available course sections
  • Need to select 5 sections for the credential
  • Possible combinations: 2,118,760
For a larger problem:
  • 200 available course sections
  • Need to select 8 sections
  • Possible combinations: 2,558,620,845,440 (2.5 trillion!)
Humans Can’t Evaluate Trade-offs Consistently: Each combination must be evaluated against:
  • 5-10 different target jobs
  • 50-200 different skills per job
  • 3 competing optimization goals (overlap, undercoverage, section count)
The genetic algorithm can evaluate thousands of combinations per second with perfect consistency. Humans would spend days or weeks on manual analysis—and likely miss the optimal solution. AI Finds Non-Obvious Solutions: The algorithm discovers course section combinations that humans wouldn’t intuitively consider:
  • Sections from different departments that complement each other
  • Non-traditional course pairings that eliminate skill gaps
  • Efficient combinations that maximize coverage with minimal sections
The genetic algorithm doesn’t replace human judgment—it augments it. The AI finds optimal combinations based on data, then you refine those results using your domain expertise and institutional knowledge.

​
Impact Analysis Deep Dive

When you toggle course sections on or off in the Curriculum tab, the system calculates impact analysis in real-time. This shows you exactly which job requirements will be affected by your change.

​
What is Impact Analysis?

Impact analysis answers the question: “If I activate/deactivate this section, which job skill requirements will cross critical thresholds?” Critical thresholds are the required proficiency levels for each skill in each job. Impact analysis identifies when your changes cause skills to:
  • Drop BELOW required levels (when deactivating sections)
  • Rise TO or ABOVE required levels (when activating sections)

​
Reading Impact Tooltips

When you hover over a section in the Curriculum tab, you see an impact tooltip that looks like this:
Example Tooltip (When Considering Deactivating a Section):Skills Dropping Below Threshold If Section Is DeactivatedSoftware Developer:
  • Python Programming (Core) - Level 4 → 2
  • SQL Databases (Relevant) - Level 3 → 1
Data Analyst:
  • Data Visualization (Core) - Level 5 → 3
  • Statistical Analysis (Relevant) - Level 4 → 2
Let’s break down each component: 1. Section Header “Skills Dropping Below Threshold If Section Is Deactivated” or “Skills Rising To Threshold If Section Is Activated” This tells you the direction of impact based on the section’s current state. 2. Job Name (e.g., “Software Developer”) Which target job’s requirements will be affected. If this section appears for multiple jobs, it means the section is critical for multiple career paths. 3. Skill Name (e.g., “Python Programming”) The specific skill that will cross a threshold. 4. Requirement Type: Core vs Relevant

Core Skills

Essential skills for job success. Employers consider these non-negotiable requirements.Impact: Failing to meet Core skill requirements significantly reduces job readiness. Students may not qualify for positions.Decision weight: HIGH - Strongly consider keeping sections that teach Core skills.

Relevant Skills

Supporting skills that enhance job performance but aren’t essential.Impact: Failing to meet Relevant skill requirements is less serious. Students may still qualify for positions.Decision weight: MEDIUM - Consider context when evaluating sections that only teach Relevant skills.
5. Level Change (e.g., “4 → 2”) Shows the proficiency level change:
  • First number (4): Current proficiency level with all active sections
  • Second number (2): New proficiency level if you toggle this section
  • Arrow direction: Down arrow (→) when deactivating, up arrow when activating
The proficiency level scale is typically 1-5 or 1-10, where higher numbers indicate greater mastery.

​
Strategic Use of Impact Analysis

Impact analysis is your primary tool for refining credentials intelligently. Here’s how to use it strategically:

​
Scenario 1: Identifying Essential Sections

1

Review Sections with Heavy Impact

Sections whose tooltips show many Core skills affected are essential. These are load-bearing sections that support multiple job requirements.
2

Mark as Must-Keep

Mentally (or literally) mark these sections as must-keep. Deactivating them would create significant coverage gaps.
3

Look for Alternatives

If an essential section has conflicts (scheduling, capacity, prerequisites), use the tooltip to understand which skills you’d need to replace. Then search for alternative sections teaching those same skills.
Example: You’re building a “Data Science Certificate” and considering whether to include “STAT 301 - Advanced Statistics (Section 02)”. Impact Tooltip shows:
  • Data Scientist: 4 Core skills drop below threshold
  • Data Analyst: 2 Core skills drop below threshold
  • Business Intelligence Analyst: 1 Relevant skill drops below threshold
Decision: This section is essential. It supports Core requirements for multiple jobs. Keep it active.

​
Scenario 2: Identifying Low-Impact Sections

1

Find Sections with Minimal Impact

Sections whose tooltips show only Relevant skills or no skills are low-impact. These are nice-to-have but not critical.
2

Consider Deactivation

If you’re trying to reduce program length or credit hours, low-impact sections are safe to remove.
3

Evaluate Trade-offs

Even if impact is low, consider other factors: Does this section provide unique value? Is it required for accreditation? Does it serve other (non-job-alignment) goals?
Example: You’re refining a “Web Development Certificate” and considering whether to include “ART 250 - Graphic Design Fundamentals”. Impact Tooltip shows:
  • Frontend Developer: 1 Relevant skill drops below threshold (Color Theory)
  • UX Designer: 1 Relevant skill drops below threshold (Visual Hierarchy)
Decision: Low job market impact. If you need to reduce credit hours, this is a safe section to remove. If you value design skills for holistic web developer education, keep it for non-alignment reasons.

​
Scenario 3: Balancing Efficiency vs Comprehensiveness

1

Identify Sections with Relevant-Only Impact

Find sections that only affect Relevant (not Core) skills across jobs.
2

Assess Credential Type

For micro-credentials (efficiency focus), consider removing these. For comprehensive certificates (mastery focus), consider keeping them.
3

Review Overall Coverage

Check the Summary tab’s skills coverage visualization. If Relevant skills are well-covered elsewhere, you can safely remove these sections. If Relevant skills are generally undercovered, consider keeping them.
Example: You’re building a “Cybersecurity Certificate” and considering “IT 490 - IT Project Management”. Impact Tooltip shows:
  • Security Analyst: 2 Relevant skills drop below threshold (Project Planning, Team Collaboration)
  • Security Engineer: 1 Relevant skill drops below threshold (Documentation)
Credential Type: Comprehensive 30-credit certificate Decision: Project management skills enhance cybersecurity professionals even if they’re not Core requirements. Keep the section for comprehensive program.

​
Scenario 4: Resolving Conflicts Between Jobs

Sometimes a section significantly impacts one job but not others.
1

Identify Job-Specific Impact

Review which jobs are affected by the section. Sometimes one job heavily relies on a section while others don’t.
2

Assess Job Priority

Are all target jobs equally important? If one job is your primary focus, prioritize sections that serve that job.
3

Consider Multi-Job Solutions

Look for sections that serve multiple jobs. These provide better overall value than sections serving only one job.
Example: You’re building a “Healthcare IT Certificate” targeting both “Clinical Informatics Specialist” and “Health Data Analyst” jobs. You’re considering “HIT 350 - Electronic Health Records Administration”. Impact Tooltip shows:
  • Clinical Informatics Specialist: 5 Core skills drop below threshold
  • Health Data Analyst: 0 skills affected
Decision: This section is critical for one job but irrelevant to the other. If both jobs are equally important, consider whether you’re designing one combined credential (include the section) or should split into two focused credentials (exclude from Data Analyst credential).

​
Advanced Impact Analysis Patterns

​
The “Replacement Strategy”

If impact analysis shows a section is critical but you can’t include it (scheduling conflicts, capacity issues, etc.), use this strategy:
1

Document Required Skills

List the Core skills that would drop below threshold if you exclude this section.
2

Search for Alternatives

Use your course catalog to find other sections teaching those same skills. Review their syllabi to confirm they cover the required skills to adequate proficiency levels.
3

Test Replacement

Deactivate the problematic section and activate the alternative. Check if impact analysis shows the skills are now adequately covered by the replacement section.
4

Validate Overall Impact

Ensure the replacement section doesn’t create new problems (excessive overlap with other sections, increases section count significantly, etc.).

​
The “Incremental Refinement Strategy”

For complex credentials with many sections, refine incrementally:
1

Start with AI Recommendations

Begin with all AI-recommended sections active (the default state after optimization).
2

Identify Lowest-Impact Section

Review impact analysis for all active sections. Find the section with the least significant impact (fewest Core skills, fewest jobs affected).
3

Deactivate and Observe

Deactivate that section. Review how the optimization score changes and whether other metrics (Total Coverage %, Curriculum Efficiency) improve or worsen.
4

Keep or Revert

If overall credential quality improves or stays the same, keep the section deactivated. If quality degrades significantly, reactivate it.
5

Repeat

Continue this process, removing the next lowest-impact section. Stop when further removals would significantly harm job readiness.
This approach helps you find the “Pareto frontier”—the point where you’ve removed all non-essential sections without compromising job readiness.

​
The “Skills Gap Analysis Strategy”

Use impact analysis to identify missing skills that no sections address:
1

Review Summary Tab

Check the embedded Skills Coverage report in the Summary tab. Identify skills with 0% coverage or coverage below required levels.
2

Search Your Catalog

Look for course sections in your catalog that might teach these missing skills (but weren’t selected by the AI).
3

Manually Activate

In the Curriculum tab, manually activate these sections. Review impact analysis to see if they fill the gaps.
4

Assess Trade-offs

Check whether manually adding sections improves overall optimization score or creates too much overlap/section count penalty. Sometimes gaps exist because filling them creates worse trade-offs.
Common Impact Analysis Pitfalls:
  1. Ignoring Relevant Skills: Don’t dismiss all Relevant skills as unimportant. In aggregate, many Relevant skills matter.
  2. Over-optimizing Section Count: Don’t remove sections just to have fewer sections if it significantly harms Core skill coverage.
  3. Toggling Without Understanding: Don’t toggle sections without reading impact tooltips. Always understand the consequences before making changes.
  4. Forgetting Non-Alignment Goals: Impact analysis only shows job market alignment. Consider other institutional goals (prerequisites, program pathways, faculty capacity) when making final decisions.

​
Best Practices for Credential Building

​
Job Selection Best Practices

Problem: Some jobs have skill requirements that your course catalog can’t address, leading to poor optimization results.Solution: Before selecting jobs, click into each job’s detail page and review its skill requirements. Verify your courses can teach those skills.Example: If a job requires “Cisco Network Certification” but your institution doesn’t offer Cisco courses, that job’s requirements can’t be met. Selecting it will pull down optimization scores.Why it works: Optimization quality depends on having courses that can teach the required skills. Selecting jobs with unmeetable requirements sets up the algorithm for failure.
Problem: Relying only on SOC jobs means missing local employer needs. Relying only on custom jobs means missing broad labor market alignment.Solution: Combine SOC jobs (national labor market) with custom jobs (local employers) for credentials that are both broadly relevant and locally aligned.Example: Select SOC job “15-1252 Software Developers” PLUS custom job “Junior Developer - [Local Employer Name]”. This ensures graduates meet both general software developer requirements and specific local employer needs.Why it works: Balanced approach provides portability (graduates can work anywhere, not just locally) and local relevance (graduates meet specific regional employer needs).
Problem: Hard to predict which job combinations will yield good optimization results until you try them.Solution: Treat job selection as iterative. Run optimization with initial jobs, review results, adjust jobs, run again.Workflow:
  1. Select initial 2-3 related jobs
  2. Run optimization
  3. Review optimization score and coverage
  4. If score is low, review which jobs have poor coverage
  5. Consider removing jobs with unmeetable requirements or adding jobs with more achievable requirements
  6. Run optimization again with adjusted jobs
Why it works: Optimization results provide feedback on job selection. Use that feedback to refine your choices.

​
Course Filtering Best Practices

Problem: Overly restrictive filters prevent the AI from finding optimal solutions.Solution: On your first optimization, don’t apply any course filters. Let the AI consider all available course sections. Only add filters if you have specific constraints.Example: Instead of immediately filtering to only “Computer Science Department” courses for a software development credential, let the AI explore. It might discover that a Statistics course or Business Analytics course provides valuable complementary skills.Why it works: The AI may find non-obvious course combinations that humans wouldn’t consider. Filters limit the search space and may exclude valuable sections.
Problem: Credentials that should align with existing academic programs need to use program courses.Solution: Use the Program filter to restrict optimization to courses within specific programs. This ensures credential-to-program alignment.Example: Creating a “Business Analytics Certificate” that should stack toward the “Business Administration A.S.” degree. Filter to only courses in the Business Administration program.Why it works: Ensures stackability and program pathway alignment. Students know the credential counts toward their degree.
Problem: Some credentials should be explicitly single-discipline (all nursing, all computer science, etc.).Solution: Use the Department filter to restrict optimization to courses within specific departments.Example: A “Clinical Nursing Skills Certificate” should only include nursing courses. Filter to Nursing Department.Why it works: Maintains disciplinary focus and meets accreditation requirements that may mandate certain courses come from specific departments.
Problem: Restrictive filters may prevent the AI from finding good solutions, resulting in low optimization scores.Solution: If optimization returns scores below 50% or shows significant skill gaps, try removing course filters and running again.Diagnostic check: If removing filters significantly improves optimization scores, your filters were too restrictive. Either keep broader search or accept that your constraints make it impossible to fully meet job requirements.Why it works: Poor results with filters often indicate the constraint is too tight. Comparing filtered vs unfiltered results shows you the cost of your constraints.

​
Optimization Settings Best Practices

Problem: Adjusting individual weight sliders without understanding how they interact leads to poor configurations.Solution: Choose one of the four configuration profiles (Maximum Job Readiness, Balanced Coverage, Focused Micro-Credential, or Exploratory) based on your credential’s primary purpose.Workflow:
  1. Review the four configuration profiles in the “Understanding Optimization Weights” section above
  2. Use the decision framework to select the appropriate profile for your credential
  3. Run optimization with your selected profile’s weight settings
  4. Review results and make minor adjustments if needed (staying within the profile’s overall approach)
Why it works: Profiles represent proven, balanced configurations that avoid conflicting priorities. They work as coordinated systems rather than independent dials.
Problem: Starting with efficiency-focused profiles (Profile 3 or 4) without first verifying your curriculum can meet job requirements.Solution: Before optimizing for brevity or efficiency, run Profile 1 (Maximum Job Readiness) to verify your course catalog has sufficient coverage.Workflow:
  1. Select your target jobs
  2. Run optimization using Profile 1 settings (Overlap 0-5%, Undercoverage 80-95%, Section Count 10-20%)
  3. Review coverage score and Skills Coverage report in Summary tab
  4. If coverage is 90%+, proceed with your desired profile
  5. If coverage is below 70%, address curriculum gaps before using other profiles
Why it works: Profile 1 reveals whether your curriculum can meet job requirements. If you can’t get good results with Profile 1, no other profile will solve the problem—you have curriculum gaps that need addressing.
Problem: Weeks or months later, you won’t remember which configuration profile you used and why.Solution: Keep notes documenting which profile you selected and your rationale.Example note:
Cybersecurity Technician Certificate
Profile: Profile 1 (Maximum Job Readiness)
Rationale: Partnership with local employers (Tech Corp, SecureNet)
           requires graduates meet 100% of Security Technician job
           requirements. Job readiness is non-negotiable.
Weights: Overlap 5%, Undercoverage 90%, Section Count 15%
Optimization Score: 87%
Notes: Initially considered Profile 2 for shorter program, but
       employer feedback emphasized need for comprehensive
       preparation. Accepted 12-course program for complete coverage.
Why it works: Documents decision-making process for institutional memory. Helps colleagues understand and replicate successful approaches.
Problem: Uncertain which profile best serves your credential goals without empirical comparison.Solution: For complex or high-stakes credentials, run optimizations with 2-3 different profiles and compare results.Comparison strategy:
  • Run Profile 1 (Maximum Job Readiness) → Note score, section count, coverage
  • Run Profile 2 (Balanced Coverage) → Note score, section count, coverage
  • Run Profile 3 (Focused Micro-Credential) if Profile 1 shows 95%+ coverage → Note score, section count, coverage
  • Review side-by-side to see trade-offs
Example:
Profile 1 Results: 88% score, 10 sections, 94% coverage
Profile 2 Results: 82% score, 7 sections, 86% coverage
Profile 3 Results: 75% score, 4 sections, 80% coverage

Decision: Profile 2 provides best balance—strong coverage (86%)
          with reasonable length (7 sections). Profile 1's extra
          3 sections add only 8% more coverage—not worth the
          additional program length for our context.
Why it works: Empirical comparison reveals actual trade-offs specific to your curriculum and jobs. Makes decision-making data-driven rather than assumption-based.

​
Refinement Best Practices

Problem: Immediately toggling sections without understanding the AI’s rationale wastes the optimization effort.Solution: Start by reviewing why the AI selected each section:
  • Read the Summary tab’s coverage analysis
  • Review the embedded Skills Coverage report
  • Understand which sections contribute to which jobs
  • Only then consider refinements
Why it works: AI recommendations are data-driven. Understanding them first helps you make informed refinements rather than undoing good decisions.
Problem: Toggling sections without understanding consequences can create unintended skill gaps.Solution: NEVER toggle a section without first hovering to read its impact tooltip. Understand which jobs and skills will be affected.Mandatory workflow:
  1. Hover over section → Read impact tooltip
  2. Understand which skills drop below/rise to thresholds
  3. Assess whether that impact is acceptable
  4. Only then toggle the section
  5. Review how optimization score changes after toggling
Why it works: Prevents blind toggling that inadvertently creates critical skill gaps. Every refinement should be informed and intentional.
Problem: Hard to find the sweet spot between comprehensive coverage and program efficiency.Solution: Use impact analysis iteratively to identify the balance point:
  1. Start with all AI-recommended sections active
  2. Deactivate the lowest-impact section (minimal Core skills affected)
  3. Check if optimization score stays acceptable and curriculum efficiency improves
  4. If yes, deactivate the next lowest-impact section
  5. If no, reactivate the last section and stop refinement
This finds the “Pareto optimal” point where you’ve removed all non-essential sections without compromising job readiness.Why it works: Systematic approach to finding the efficiency-comprehensiveness balance using objective impact data.
Problem: First refinement attempt may not be the best possible credential.Solution: After initial refinement, use the “Reset” button to restore AI recommendations, then try different refinement strategies.Strategies to test:
  • Scenario A: Minimize section count (deactivate all sections with only Relevant skill impacts)
  • Scenario B: Maximize Core skill coverage (keep all sections affecting Core skills)
  • Scenario C: Custom based on your institution’s priorities (scheduling, prerequisites, etc.)
Compare scenarios to see which best serves your goals.Why it works: Refinement is creative problem-solving. Testing multiple approaches helps you find the best solution.

​
Credential Creation Best Practices

Problem: Generic names like “Badge 1” or “Certificate A” aren’t meaningful to students or administrators.Solution: Create clear, specific credential names that describe both the content and target careers.Good examples:
  • “Cybersecurity Fundamentals Badge”
  • “Full-Stack Web Development Certificate”
  • “Clinical Nursing Skills Micro-Credential”
  • “Data Analytics for Business Professionals Certificate”
Bad examples:
  • “Certificate 1”
  • “IT Badge”
  • “Stackable Credential”
Why it works: Descriptive names help students understand what the credential offers and administrators track multiple credentials clearly.
Problem: Students and advisors need to understand the credential’s purpose, target careers, and outcomes.Solution: Write comprehensive credential descriptions that include:
  • Purpose: What is this credential for?
  • Target careers: Which jobs does it prepare students for?
  • Student outcomes: What will students be able to do after completing this credential?
  • Stackability: How does this credential relate to other credentials or degree programs?
  • Estimated completion time: How long will this credential take to complete?
Example description:
The Cybersecurity Fundamentals Badge prepares students for entry-level
cybersecurity roles including Security Analyst and IT Security Specialist.

Students will gain hands-on skills in network security, threat detection,
cryptography fundamentals, and security incident response.

This 12-credit badge stacks toward the Cybersecurity A.A.S. degree and can
be completed in one semester while attending full-time.
Why it works: Clear descriptions help students make informed enrollment decisions and advisors recommend appropriate credentials.
Problem: Months or years later, you may forget which jobs the credential was optimized for.Solution: Include target job information in the credential description or notes. List the specific jobs (including SOC codes if applicable) the credential prepares students for.Example:
Target Career Preparation:
- Software Developer (SOC 15-1252)
- Web Developer (SOC 15-1254)
- Junior Developer at [Local Employer Partners: Company A, Company B]
Why it works: Maintains institutional knowledge about credential design rationale. Helps with future updates and quality assurance reviews.
Problem: Once created, credentials may be difficult to modify, especially if students are already enrolled.Solution: Before clicking “Create Credential”:
  1. Double-check all active sections are correct
  2. Review optimization score is acceptable
  3. Verify skills coverage in Summary tab
  4. Confirm credential name and description are polished
  5. Have a colleague review if high-stakes credential
Only create the credential once you’re confident in every detail.Why it works: Prevents having to modify credentials after student enrollment begins. Ensures credential quality from day one.

​
When to Start Over

Sometimes it’s better to reset and start fresh rather than continuing to refine poor results. Reset and start over if:
  • Optimization score consistently below 50%: Indicates fundamental mismatch between job requirements and available courses. Refining won’t fix this.
  • No sections selected: Algorithm couldn’t find any combinations meeting your constraints. Filters are too restrictive or job requirements aren’t teachable with available courses.
  • Results don’t match expectations: If optimized curriculum is completely unexpected (wrong courses, wrong level, etc.), constraints may be misconfigured.
  • Conflicting goals identified: If you realize your weight settings create impossible-to-satisfy conflicts (e.g., all three penalties at 100%).
When you reset, try:
1

Simpler Configuration

Reduce complexity: fewer jobs (2-3 instead of 6-7), no course filters, default weights.
2

Different Job Selection

If initial jobs had poor coverage, try different related jobs or check that your selected jobs have realistic skill requirements.
3

Verify Data Quality

Confirm your course sections have processed syllabi with skills data. Confirm your jobs have detailed skill requirements. Poor input data causes poor optimization results.
4

Adjust Expectations

If optimization consistently can’t meet requirements, accept that your current course catalog may have gaps. Consider whether you need to develop new courses or adjust target jobs.

​
Common Pitfalls to Avoid

Pitfall 1: Selecting Too Many Diverse JobsProblem: Credential tries to prepare students for 8-10 unrelated careers. Algorithm can’t find sections meeting all diverse requirements.Solution: Focus on 2-4 related jobs from the same career cluster. Create multiple focused credentials rather than one unfocused mega-credential.
Pitfall 2: All Three Weights at Extreme ValuesProblem: Setting overlap penalty, undercoverage penalty, AND section count penalty all to 80-100% creates impossible constraints.Solution: Emphasize 1-2 weights, keep the third moderate. Accept that you can’t simultaneously maximize coverage, minimize overlap, AND minimize section count.
Pitfall 3: Over-Toggling Sections Without UnderstandingProblem: Rapidly toggling many sections without reading impact analysis. Creates unforeseen skill gaps and poor job alignment.Solution: Slow down. Read impact tooltips. Toggle one section at a time. Observe how each change affects optimization score and coverage.
Pitfall 4: Ignoring Impact TooltipsProblem: Toggling sections blindly, ignoring impact analysis warnings that Core skills are dropping below thresholds.Solution: Make impact analysis review mandatory before every toggle. Treat warnings about Core skills seriously.
Pitfall 5: Creating Credentials Without Reviewing ResultsProblem: Running optimization, seeing a score, and immediately creating credential without reviewing curriculum or coverage details.Solution: Always review Summary tab skills coverage, Curriculum tab section list, and embedded Skills Coverage report before creating credentials. Understand what you’re creating.
Pitfall 6: Forgetting About Non-Alignment CriteriaProblem: Optimization only considers job market alignment. Forgetting about prerequisites, scheduling, faculty capacity, accreditation requirements, etc.Solution: After optimization, review optimized curriculum through additional institutional lenses:
  • Can students actually schedule these sections?
  • Are prerequisites manageable?
  • Do we have faculty capacity to offer these sections?
  • Do accreditation/program requirements mandate certain courses?
Optimization provides the job-aligned foundation; you layer institutional reality on top.

​
Advanced Troubleshooting

​
Issue: Optimization Always Returns Same Sections

Symptoms:
  • Running optimization multiple times with different settings
  • Algorithm consistently selects the same 4-5 sections
  • Changing weights doesn’t significantly alter results
Possible Causes:
  1. Limited course options: Very few sections teach the required skills, so AI has limited choices
  2. Filters too restrictive: Course filters are limiting available sections to a small pool
  3. Dominant sections: A few sections teach many critical skills, making them always optimal choices
Diagnosis:
  • Remove all course filters and run optimization again. If results change dramatically, filters were too restrictive.
  • Review target jobs’ skill requirements. Count how many course sections teach each required skill.
  • Check the Summary tab’s Skills Coverage report. Look for skills with only 1-2 contributing sections.
Solutions:
1

Remove or Loosen Filters

If filters are causing the limitation, remove them or expand them (add more programs/departments).
2

Add More Course Variety

If your catalog simply lacks diversity in certain skill areas, consider developing new courses or adding more sections with varied content.
3

Accept the Result

If the same sections consistently appear because they’re objectively the best options (teach many critical skills well), accept that the algorithm is working correctly. Those sections are genuinely optimal.
4

Try Different Target Jobs

If you want different sections, consider whether you’re targeting the right jobs. Different jobs have different skill requirements that would favor different sections.

​
Issue: Optimization Scores Won’t Exceed 70% Despite Adjustments

Symptoms:
  • Optimization score stuck around 65-70%
  • Adjusting weights, jobs, filters doesn’t improve score beyond 70%
  • Skills Coverage report shows persistent gaps
Possible Causes:
  1. Job requirements exceed course capabilities: Target jobs require skills at proficiency levels your courses can’t teach
  2. Missing skills: Jobs require skills that no courses in your catalog teach
  3. Unrealistic job selection: Selected jobs require skills from different domains that can’t be covered in a single reasonable credential
Diagnosis:
  • Go to Summary tab → Review embedded Skills Coverage report
  • Identify skills with 0% coverage or coverage significantly below required levels
  • Click into your target jobs’ detail pages → Review their skill requirements → Check which required skills have no courses
  • Compare skill requirements across your selected jobs → Check if jobs share common skills or require completely different skills
Solutions:
1

Review Specific Gaps

Use the Skills Coverage report to identify exactly which skills are missing or undercovered. Make a list.
2

Check Your Course Catalog

For each missing skill, search your course catalog. Do any courses teach this skill but weren’t selected? If so, manually activate them and see if coverage improves.
3

Assess Catalog Limitations

If skills genuinely aren’t taught anywhere in your catalog, you have a curriculum gap. Optimization can’t fix this—you need to develop new courses or sections.
4

Adjust Job Selection

If jobs require skills from vastly different domains (e.g., nursing + programming + accounting), consider whether you’re trying to create one credential that should actually be three separate credentials.
5

Accept Realistic Scores

70% may be the best achievable score given your current catalog and target jobs. Not all credentials can reach 90-100% if the job requirements are extensive and your courses have limited coverage.
6

Contact Support for Gap Analysis

For complex situations where you’re unsure why scores are limited, contact support@mapademics.com. We can provide detailed curriculum gap analysis showing exactly which courses you’d need to develop to improve coverage.

​
Issue: Using Wrong Configuration Profile for Credential Goal

Symptoms:
  • Optimization results don’t match expectations
  • Getting short credentials when you need comprehensive job preparation
  • Getting low scores despite having relevant courses
  • System seems to skip teaching required skills
Common Mistakes:
  1. Using Profile 3 (Focused Micro-Credential) as starting point: Trying to optimize for brevity before verifying curriculum can meet job requirements
  2. Using Profile 4 (Exploratory) for job-readiness credentials: Accepting gaps when students need complete preparation
  3. Using Profile 1 (Maximum Job Readiness) for introductory badges: Expecting complete coverage for credentials that should be foundational only
Diagnosis:
  • Review your credential’s purpose: Is this for job readiness, career exploration, or something else?
  • Check which profile you selected: Does it match your stated goal?
  • Compare results across profiles: Run optimization with 2-3 different profiles to see the difference
Solutions:
1

Clarify Credential Purpose

Define whether this credential should:
  • Prepare students for immediate employment (use Profile 1 or 2)
  • Introduce students to a field (use Profile 4)
  • Reduce redundancy in already-comprehensive curriculum (use Profile 3 only after Profile 1 verification)
2

Match Profile to Purpose

Use the decision framework in the “How to Choose the Right Profile” section to select the appropriate profile:
  • Job readiness required → Profile 1
  • Balanced approach → Profile 2
  • Already verified 95%+ coverage → Profile 3
  • Introductory/exploratory → Profile 4
3

Verify with Profile 1 First

Before using Profile 3 or 4, always run Profile 1 to verify your curriculum can meet job requirements. If Profile 1 shows low scores, you have curriculum gaps—no other profile will solve this.
4

Re-run with Correct Profile

Reset optimization and run again with the profile that matches your credential’s actual purpose.
Example: 
Problem: Using Profile 3 for a Registered Nurse credential
- Profile 3 settings: Overlap 35%, Undercoverage 55%, Section Count 60%
- Result: 4-course credential with 68% coverage
- Issue: Nursing requires 100% coverage—Profile 3 accepts gaps

Solution: Switch to Profile 1
- Profile 1 settings: Overlap 5%, Undercoverage 95%, Section Count 15%
- Result: 12-course credential with 98% coverage
- Outcome: Comprehensive preparation meeting accreditation requirements

​
Issue: Want to Save Optimization Configuration for Reuse

Symptoms:
  • Creating multiple similar credentials (e.g., several micro-credentials in the same field)
  • Want to reuse successful job selection, filter, and weight configurations
  • Current system doesn’t save configurations between sessions
Current Limitation: The Credential Builder currently doesn’t save optimization configurations. Each new credential design starts from scratch. Workarounds:
1

Document Configurations Externally

Keep a spreadsheet or document tracking successful configurations:
  • Credential name/type
  • Target jobs (with SOC codes/IDs)
  • Course filters (programs, departments, courses)
  • Weight settings (overlap %, undercoverage %, section count %)
  • Resulting optimization score
  • Notes on why this configuration worked
2

Create Reusable Templates

For credentials you create frequently (e.g., monthly workforce training badges), create template documents with:
  • Standard job selection for that credential type
  • Standard filter settings
  • Standard weight settings
  • Step-by-step instructions for using this template
3

Share Best Configurations With Team

If multiple team members design credentials, share successful configurations so everyone can benefit from discovered best practices.
Future Feature: Configuration templates and saved presets are on the product roadmap. You’ll eventually be able to save, name, and reuse optimization configurations. Contact support@mapademics.com to express interest in this feature.

​
Issue: Need Multiple Credentials for Same Jobs With Different Focuses

Scenario: Creating multiple credentials targeting the same careers but emphasizing different aspects:
  • A quick, efficient micro-credential
  • A comprehensive certificate
  • A specialized advanced credential
Strategy: Use optimization weights to create different emphases from the same job selection:

Micro-Credential Version

Jobs: Same target careersWeights:
  • Overlap: 85% (high)
  • Undercoverage: 70% (medium-high)
  • Section Count: 85% (high)
Result: Short, efficient credential covering core skills. Accepts some gaps for brevity.

Certificate Version

Jobs: Same target careersWeights:
  • Overlap: 30% (low)
  • Undercoverage: 90% (very high)
  • Section Count: 30% (low)
Result: Comprehensive credential covering all skills thoroughly. Longer program but complete preparation.

Advanced Specialization

Jobs: Same target careers + advanced variationsWeights:
  • Overlap: 50% (medium)
  • Undercoverage: 85% (high)
  • Section Count: 40% (medium)
Filters: Limit to upper-level courses (300-400 level)Result: Advanced credential assuming foundational knowledge. Focuses on specialized, advanced skills.
Workflow:
  1. Design micro-credential first (efficiency focus)
  2. Design comprehensive certificate second (coverage focus)
  3. Ensure micro-credential sections are subset of certificate sections (stackability)
  4. Design advanced specialization third, filtering to exclude intro sections
This creates a credential pathway where students can start with micro-credential, continue to certificate, then pursue advanced specialization.

​
Issue: Optimization Takes Longer Than Expected (>5 Minutes)

Symptoms:
  • Optimization running for 5-10+ minutes
  • Progress stuck on same generation for extended time
  • Browser shows “Waiting for optimization…” indefinitely
Possible Causes:
  1. Too many target jobs: Selected 10+ jobs, creating enormous search space
  2. Very large course catalog: 500+ sections available, exploding combinations
  3. Server overload: Multiple users running optimizations simultaneously
  4. Network/WebSocket disconnection: Real-time updates not reaching your browser
Solutions:
1

Reduce Target Jobs

If you selected 8+ jobs, cancel optimization and reduce to 3-5 jobs. More jobs dramatically increase computation time.
2

Apply Course Filters

If your catalog has 200+ sections, apply filters to reduce the search space:
  • Filter to specific programs or departments
  • Filter to specific course level ranges
  • Exclude courses you know aren’t relevant
This reduces combinations the algorithm must evaluate.
3

Check WebSocket Connection

Look for the real-time connection status indicator (should show “Connected”). If disconnected, refresh the page and try again.
4

Wait for Server Queue

If running during peak usage times, your optimization may be queued. Check the Background Jobs page to see if your optimization is queued or running.
5

Cancel and Retry

If optimization exceeds 10 minutes without progress, cancel it. Simplify your configuration (fewer jobs, more filters) and run again.
6

Contact Support

If simplified configurations still take excessive time, contact support@mapademics.com. There may be a system performance issue.

​
Issue: Need to Design Credentials for Multi-Track Programs

Scenario: Creating credentials for programs with multiple specialization tracks (e.g., Computer Science with Web Development and Data Science tracks). Strategy:
1

Design Core Credential First

Create a foundational credential covering skills common to all tracks:
  • Select jobs representing all tracks
  • Use high undercoverage penalty to ensure foundational coverage
  • Create credential named “[Program Name] Core Foundations”
2

Design Track-Specific Credentials

Create separate credentials for each track:
  • Select jobs specific to that track
  • Apply course filters excluding core foundation courses (to avoid overlap)
  • Use moderate overlap penalty (some reinforcement of core skills in new contexts is okay)
  • Create credentials named “[Program Name] - [Track Name] Specialization”
3

Document Stackability

In credential descriptions, explicitly state:
  • Core credential is prerequisite for specializations
  • Specializations stack with core to form complete track
  • Students can pursue multiple specializations if desired
4

Verify Coverage

Create a test report including both core and specialization credentials. Verify that combined credentials provide comprehensive coverage of track jobs.
Example: 
Computer Science Program Structure:

Core Credential (12 credits):
- "Computer Science Foundations Badge"
- Covers programming fundamentals, algorithms, data structures
- Required for all tracks

Specialization Credentials (9-12 credits each):
- "Web Development Specialization Certificate"
- "Data Science Specialization Certificate"
- "Cybersecurity Specialization Certificate"

Students complete Core + one (or more) Specializations

​
Getting Expert Support

For complex credential design scenarios, curriculum gap analysis, or strategic optimization consultation, Mapademics support can help. When to contact support:
  • Optimization consistently produces poor results despite troubleshooting
  • Need curriculum gap analysis showing which courses to develop
  • Designing complex multi-track credential pathways
  • Need help interpreting Skills Coverage reports
  • Want strategic guidance on credential portfolio design
  • Technical issues with optimization (errors, timeouts, unexpected behavior)
Contact: support@mapademics.com Include in your message:
  • Brief description of your credential design goals
  • Target jobs you’re trying to prepare students for
  • Specific challenges you’re encountering
  • Screenshots of optimization results if relevant
  • Any error messages or unexpected behaviors
Support can provide:
  • Detailed curriculum gap analysis
  • Recommendations for weight configurations
  • Guidance on job selection strategy
  • Troubleshooting for technical issues
  • Strategic consultation on credential design

​
Complete Credential Builder Documentation

Introduction

Learn what the Credential Builder is, when to use it, and how it works at a high level.

Building Credentials

Complete step-by-step guide to using the Credential Builder from start to finish.

Advanced Topics

You are here: Deep dive into optimization weights, algorithm mechanics, and advanced strategies.

​
Additional Resources

Managing Stackable Credentials

Learn how to manage credentials after creation, add/remove sections, and track student progress.

Skills Processing Overview

Understand how Mapademics extracts skills from syllabi and job descriptions to power the Credential Builder.

Report Designer

Create custom reports analyzing credential effectiveness and skills alignment across programs.

Background Jobs Monitoring

Monitor optimization jobs and other background processing tasks in your organization.