Coursera
EdTech · AI Feature Design · Live Prototype

Coursera:
Adaptive Assignment Coach

Designing an in-context AI tutoring system that transforms the "I'm stuck" moment from a point of abandonment into a deeper learning opportunity — without giving away the answer.

TypeAI Feature Addition
DomainEdTech / STEM
DeliverablesPRD + Prototype + Deck
StatusPrototype Live

The Problem

Learners in technical courses hit a wall during graded coding assignments. Coursera's current support options — forum posts and static hints — are asynchronous, generic, and slow. When a learner gets stuck at 11pm on a data structures problem, there's no one to help them.

This creates two failure modes:

Failure Mode 1

Abandonment: Learner closes the tab, loses momentum, and eventually churns from the course entirely.

Failure Mode 2

Integrity Violation: Learner copies a solution from GitHub, Chegg, or Stack Overflow — completing the assignment without learning anything.

Both outcomes hurt Coursera's core mission: providing universal access to world-class learning. And both are preventable if the right kind of help is available at the right moment.

P0 Problem Statement

Learners in technical STEM courses reach a "Cognitive Wall" during graded assignments. Current support is asynchronous and non-personalized, leading to course abandonment or academic dishonesty — at the exact moment learners are closest to a breakthrough.

Market Context

The problem is concentrated in STEM and professional data/development specializations — the highest-value, highest-dropout-rate courses on Coursera's platform.

2.5×Target increase in assignment pass rates
100M+Coursera learners globally
300+University & company partners

Competitive Analysis

CompetitorApproachGap
Codecademy "Get Unstuck"Hint systemOften reveals full solution — reduces cognitive load too much, no learning
Khan Academy (Khanmigo)Socratic AI tutorStrong model but no deep integration with Coursera's grading infrastructure or instructor solution graphs
Generic LLM wrappersChatGPT-style chatWill generate complete solutions on request — destroys assessment integrity
Adaptive Assignment CoachSocratic + AST comparisonCompares learner's code to instructor's solution graph — provides precise nudges, never answers

The key differentiation: our coach doesn't just "guess" based on the prompt. It ingests the instructor's solution and converts it into a logic flow, then maps the learner's attempt against it to pinpoint exactly where reasoning diverged.

User Personas

Priya — The Career Switcher
29 · QA Engineer → Data Automation

Studies around a full-time job. Understands videos conceptually but freezes at the blank code editor.

"I understand the video, but the blank code editor is terrifying."

Mark — The Upskiller
32 · Developer updating Python skills

Needs efficient debugging help. Gets the logic right but loses hours on syntax errors he can't diagnose.

"I have the logic right, but I've been stuck on a NoneType error for two hours."

Secondary: Course Author

Course Author Needs

"I can't manually hint 10,000 students, but I don't want AI to give them the answers." The coach must scale their intent — not replace their pedagogy.

The Solution: Socratic Logic Engine

A "Need Help?" floating action button triggers an AI tutoring side panel that lives inside the Coursera Lab interface — non-intrusive, contextual, and hard-coded to never generate complete solutions.

Four Help Modes

ModeWhat It DoesWhen Used
Plain LanguageRephrases the assignment prompt in simple termsLearner isn't sure what they're being asked to build
Similar Pattern"Brother Problems" — shows a structurally similar solved exampleLearner understands the task but doesn't know where to start
Approach AnalysisMaps learner's attempt line-by-line against instructor's solution path, identifies exact divergence pointLearner has written code but it's wrong
Practice ModeGenerates 3 variant problems with different variables to cement the conceptAfter passing — verify real understanding, not lucky guesses
The Core Guardrail (FR.1)

The system is hard-coded to strip any "Full Solution" strings from its output. It provides scaffolding and hints only. This is non-negotiable — the moment it generates answers, the product destroys its own value proposition.

Technical Architecture

Canonical Solution Graph: Course authors provide their solution. The system ingests it and converts it into a logic flow (Abstract Syntax Tree). Every learner hint is grounded in this — not in the LLM's general knowledge.

State Management: The coach maintains full history of the current assignment session to avoid repetitive hints and track learner progress toward the solution.

Model Strategy: Gemini 1.5 Flash for speed on simple task clarification. GPT-4o for complex code analysis. Rate limited to 5 hints per assignment per hour to prevent brute-forcing.

Privacy: User-pasted code is anonymized before being sent to any LLM API.

Grader Integration: Hooks into Coursera's Judge0 (or equivalent) grader so the coach can see submission history and tailor hints accordingly.

User Stories

IDUser StoryAcceptance Criteria
US.1As a learner, I want the coach to rephrase the prompt so I can verify I haven't misunderstood the taskPlain-language restatement appears within 3 seconds of clicking "Plain Language" mode
US.2As a learner, I want to know where my logic failed without seeing the answer written for meSystem identifies divergence point (e.g., "Line 14 handles the loop incorrectly") — no corrected code provided
US.3As a learner, I want similar example problems to map concepts to new contexts3 structurally similar problems generated with distinct variables/scenarios, all auto-gradeable
US.4As a course author, I want the AI to never reveal my solutionZero full solution strings in any API output — enforced at system prompt and output filter layer

Success Metrics

15%Target increase in assignment completion rates (pilot)
<3AI interactions to pass (Hint Efficacy KPI)
24hStuck-to-Submit rate window
MetricDefinitionWhy It Matters
Stuck-to-Submit Rate% of users who submit within 24h of clicking "I'm Stuck"Primary signal that the coach is actually unblocking learners
Hint Efficacy% who pass after <3 AI interactionsValidates that hints are genuinely helpful, not just increasing usage
Practice Set Conversion% who engage with "Generate 3 Similar Problems"Signals genuine learning intent vs. just wanting to pass
Hint Abuse Rate% attempting to extract solutions via prompt manipulationGuardrail effectiveness monitoring

Rollout Strategy

PhaseScopeGoal
Alpha (Wk 1–4)Internal testing · "Python for Everybody" onlyValidate guardrails and hint quality
Beta (Wk 5–9)A/B test on 10% of learners in 3 high-traffic STEM specializationsMeasure Stuck-to-Submit lift vs. control
GA (Wk 12+)All auto-graded coding and math assignmentsFull rollout based on beta signal

Risks & Mitigations

RiskImpactMitigation
HallucinationHighGround AI exclusively in Canonical Solution — restrict knowledge to specific assignment parameters only
Hint AbuseMedium"Hint Credit" system — 5 hints per assignment per hour with cooldown timer
Model LatencyLowFlash model for simple clarification, larger model only for code analysis
View Live Prototype ↗