Inspiration
Traditional learning platforms often measure whether a learner selected the correct answer, but they do not provide a deeper understanding of why the learner is struggling or whether the learner can actually apply the concept.
Generative AI can already explain concepts and generate questions. We wanted to build something beyond a simple AI quiz generator — a system that uses AI to measure learner competency through evidence, identify root skill gaps, and guide the learner toward targeted improvement.
This led us to build SkillLens AI, an AI-powered competency intelligence platform for learning.
What We Built
SkillLens AI converts learning material into a structured competency map and uses it to create personalized assessments.
The platform follows a continuous learning cycle:
Learning Material → Competency Mapping → Assessment → Evidence → Competency Analysis → Skill Gap Detection → Personalized Learning → Re-assessment
Instead of giving every learner the same assessment, the system analyzes performance at the topic and competency level.
For example, a learner may score well in Java Classes but struggle with Inheritance and Polymorphism. Rather than simply reporting a low score, SkillLens AI identifies the underlying gap and recommends what the learner should study next.
How It Works
- A teacher or administrator uploads learning material such as course notes or documents.
- The system processes the material and identifies important topics and competencies.
- AI generates assessment questions based on the provided learning content.
- Learners attempt the assessment.
- The platform analyzes performance at the individual competency level.
- Weak areas are identified using assessment evidence and historical performance.
- AI generates personalized recommendations and targeted practice.
- Learners can take follow-up assessments to measure improvement.
Our long-term goal is to create a Competency Passport where a learner's skills are represented not only by scores, but also by assessment evidence, confidence, progress, and demonstrated improvement.
What Makes It Different
The core idea is not simply to generate AI questions.
ChatGPT can generate an answer. SkillLens AI focuses on measuring whether a learner can demonstrate the competency behind that answer.
The platform combines AI generation with structured competency mapping, evidence-based assessment, adaptive learning, root-cause skill-gap analysis, and continuous re-assessment.
This creates a learning loop instead of a one-time quiz.
Challenges
One of the biggest challenges was designing the system so that AI remains useful without making the entire assessment process a black box.
We therefore designed the platform around structured data and evidence. AI is used for tasks such as content understanding, question generation, competency analysis, and personalized recommendations, while the platform maintains the learner's assessment history, competency scores, and evidence.
Another challenge was keeping the architecture practical enough for a working prototype. Instead of training our own large language model, we designed an AI integration layer that can work with existing LLM APIs while keeping the competency and assessment logic within our platform.
Future Scope
SkillLens AI can be extended with:
- Adaptive difficulty based on learner performance
- Coding and debugging assessments
- AI-based practical competency evaluation
- Teacher and institution-level analytics
- Competency confidence scoring
- Competency Passport and verified skill history
- RAG-based source-grounded learning
- Multi-course and cross-domain competency graphs
Our vision is to move learning analytics from "What score did the learner get?" to "What can the learner actually do, what are they struggling with, and what should they learn next?"
Built With
- ai
- api
- auth
- css
- generative
- github
- llm
- next.js
- postgresql
- react
- recharts
- rest
- storage
- supabase
- tailwind
- typescript
