AcuLearn — AI-Powered Course Advisor

Context

Jan 2025 · Completed competition prototype · Maturity: Prototype.

Problem

A static course catalog could not adapt to a learner’s existing knowledge, goals, or gaps. The experience needed to recommend only the relevant parts of a learning path.

Users

Prospective product-management learners in a Bozhan product competition.

My Role

Two-person product team; product management, conversation/prompt design, Telegram + AI integration, with technical implementation by Alireza.

Responsibilities

• Defined the product problem, value proposition, conversation UX, assessment criteria, recommendation logic, and learning-sequence behavior.
• Designed contextual continuity and a human-support fallback for cases the AI could not resolve confidently.
• Collaborated with the second product teammate on questions and evaluation criteria while owning the main technical flow.

Constraints

• Competition timeframe and prototype maturity.
• The AI interview implementation lived partly in a separate repository.
• No commercial-scale or production-traction claim.

Product Decisions

• Interview for both current knowledge and learning goal before recommending.
• Recommend course modules and sequencing rather than treat the experience as a quiz.
• Use a human fallback when the product cannot resolve the case safely.

Delivery & Technical Approach

• Implemented the Telegram and interview flows with PHP, webhooks, command routing, deep links, and MySQL/Medoo state.
• Integrated GPT through Metis, maintained conversation context, and connected the AI flow to a knowledge-base/RAG architecture recorded in the canonical source.
• Connected unresolved cases to human support.

Product flow

1. Telegram entry
2. Dynamic AI interview
3. Knowledge & goal assessment
4. Course-module selection
5. Personalized sequence
6. Context persistence
7. Human-support fallback

Evidence / Outcome

• Approximately 500 users started the flow.
• Approximately 300 completed the full intelligent interview.
• Placed third in a product competition organized by Bozhan.

What I Learned

• Conversation design, recommendation logic, and fallback behavior matter as much as the underlying model.
• Completion metrics are only useful when the flow being completed is defined precisely.

Evidence Boundary

Next Iteration

Validate recommendation usefulness after the interview, not only interview completion, and strengthen retrieval/evaluation evidence for the knowledge-base layer.