Digital Twin
AI Pipeline for Student Academic Insights
Overview
An AI pipeline that turns a student's academic data into personalized insights, advice, and recommendations — helping university students see where they stand and what to do next. v1 in progress.
- Building an LLM pipeline (Google Gemini) that turns structured student academic data into tailored insights and recommendations.
- I own the core pipeline end-to-end — the request → context → model → response flow.
- Collaborative build with batchmates; v1 still in progress.
The problem
University students rarely get personalized, data-driven guidance — generic advice ignores an individual's actual academic record.
My approach
An AI pipeline that ingests a student's structured academic data and uses an LLM (Gemini) to generate tailored insights, advice, and recommendations.
What I owned
I own the core AI pipeline: the request → context → model → response flow that turns raw academic data into a grounded prompt and a usable response.
The hard part
The core challenge I'm working through is grounding the LLM in each student's real academic data, so the insights stay specific and accurate instead of generic advice.
v1 is still in progress — I'm keeping this honest to where it stands today, and I'll expand the write-up once v1 ships.