Junction Education is an edtech platform delivering personalized learning experiences for students. Partnering with GoML, the company developed an AI-powered system that enables learning objective-based education, generating dynamic study materials and interactive assessments aligned with specific learning objectives and student performance.
Problem: Scaling personalized learning and assessment
Junction Education wanted to introduce a learning objective–based approach for a platform serving 50,000 to 100,000 active users. The aim was to move away from one-size-fits-all content and deliver personalized learning experiences that adapt to each student’s performance and specific learning goals.
To do this, the platform needed to generate objective-aligned study materials, provide real-time feedback and adaptive assessments, dynamically integrate multimedia content, and track student progress to adjust difficulty levels over time. All of this had to scale to tens of thousands of concurrent learners, requiring an AI-powered system that could produce accurate, personalized content in seconds rather than hours.
The solution: AI-powered personalized learning platform
GoML designed and delivered a 9-week MVP to enable learning objective-based education on Junction Education's platform. The solution generates personalized study materials, interactive quizzes, and multimedia resources dynamically based on specific learning objectives and individual student performance.
An agentic AI orchestration layer manages workflow, coordinating interactions between specialized agents. Diagnostic Chat Agent, Content Generation Agent, and Quiz Agent and triggering tools for image and video mapping.
Key goals of the solution include:
- AI content generation: Produces structured, personalized study materials and simplified summaries aligned with student performance and learning objectives.
- Interactive assessments: Generates MCQs and fill-in-the-blank quizzes with real-time feedback via a chat interface.
- Multimedia integration: Maps existing images and videos to content and dynamically generates diagrams when needed.
- Agentic orchestration: Coordinates Bedrock Agents and Lambda Action Groups for scalable, context-aware AI workflows.
By combining AI content generation with agentic orchestration, Junction Education delivers a personalized learning platform that is secure, scalable, and interactive, reducing manual effort while enhancing student engagement.
1. Intelligent content generation
AI generates personalized study materials by summarizing, simplifying, and structuring educational content according to learning objectives and student performance.
2. Video mapping and generation
Videos are automatically mapped to relevant learning objectives and student queries, providing targeted multimedia support within the personalized learning platform.
3. AI-Based image generation and mapping
- Relevant diagrams, charts, and illustrations are generated or retrieved automatically.
- Claude via AWS Bedrock creates SVG-based visuals when no existing images are available.
- Metadata tagging ensures fast retrieval from the Knowledge Base, enhancing the personalized learning platform experience.
4. Interactive quiz generation
- AI generates MCQs and fill-in-the-blank quizzes in real time.
- Quizzes provide immediate feedback, track attempts, and automatically deactivate after completion or max attempts, supporting adaptive learning in the personalized learning platform.
5. Conversational interface
- WebSocket-based chat interface allows students to interact with study materials, quizzes, and AI tutors.
- Diagnostic Chat Agent clarifies concepts and guides learning.
- Quiz Agent manages interactive assessments and structured feedback, making the personalized learning platform engaging and responsive.
6. Backend infrastructure
- AWS-based architecture with Lambda, S3, and Bedrock integration.
- Pinecone vector database supports fast retrieval of content embeddings.
- MongoDB and Redis manage session data, conversation history, and quiz tracking.
7. Testing and validation
- End-to-end testing validates chat flows, content accuracy, and quiz logic.
- Performance tests ensure high concurrency and smooth content streaming.
- Accessibility checks align with WCAG 2.2 standards.
Impact of the AI-powered personalized learning platform
- 70–80% reduction in manual content creation for educators.
- Real-time feedback improves student engagement and understanding.
- Integrated multimedia resources enrich learning experiences.
- Centralized AI platform enables scalable, self-service content and assessment workflows.
About
Before AI vs after AI
“With Junction Education’s personalized learning platform, we transformed content delivery from static and manual to dynamic, interactive, and tailored for each student. Students receive instant feedback, educators can scale content efficiently, and the system integrates multimedia resources seamlessly.” - Prashanna Rao, Head of Engineering, GoML
Key takeaways for education platforms
Common challenges
- Manual content creation slows course updates.
- Limited interactivity reduces engagement.
- No centralized mechanism for student performance tracking.
Practical guidance
- Use AI content generation to produce study materials aligned with learning objectives.
- Integrate interactive quizzes with immediate feedback.
- Automatically map multimedia resources to content topics.
- Build a backend that supports scalability, secure access, and agent orchestration for a personalized learning platform.
Ready to modernize your learning platform?
Partner with GoML to build a personalized learning platform that delivers AI-generated study content, interactive assessments, and multimedia integration at scale.
