// Project 01 — AI Singapore

AI-Powered Student Success Platform

A decision-support platform that helps lecturers identify students needing additional support and generate personalised intervention recommendations. Analytics, conversational AI and recommendation generation sit behind a natural language interface, so lecturers ask questions directly instead of reading across multiple reports.

Amazon Bedrock AgentCore
Python
AWS S3
LLM Agents

Mockup of the chat interface. A lecturer asks 'how are my classes doing?' and the assistant returns a table of three classes with average mark, attendance and pass rate, ranked weakest first, followed by a short note on each class, suggested follow-up questions, and a chip naming the term and the date of the latest data. — Mockup with synthetic data — the lecturer asks in plain language; the answer arrives with follow-up questions and the date of the data it was computed from.

// Engineering highlights

Conversational layer. Built on Amazon Bedrock AgentCore, with long-term memory carrying context across sessions and modular tool interfaces for retrieving analytics and recommendation outputs.
Never reading half-written data. Agents risked resolving partially published datasets. I added a pointer-based commit mechanism using lightweight JSON pointers in S3 — pointers update in a strict sequence and the final state pointer acts as the commit signal, so analytics only ever run on complete, consistent data.
Numbers the model can't invent. Evaluation showed the LLM occasionally fabricating statistics rather than reporting computed values. Moving all numerical aggregation into deterministic Python tools left the model responsible only for presenting verified results in natural language.

// Implementation details limited by confidentiality.

Next project Word World — AI Dictionary for Kids