Lumen An AI study companion
Placeholder. A retrieval-augmented assistant that answers questions from your own course notes — with citations, so you can check its work.
¹/ aa·rav sen /noun
Move across the map
Each region is a different part of how I work. The overlap is where I’m heading.
I started out learning to write code. I stayed because I realised code is only half the story — the other half is understanding which problems are worth solving, and why.
Placeholder paragraph. This is where a short, honest introduction goes — what you study, what pulled you towards AI, and the moment you became curious about how businesses actually work.
Placeholder paragraph. A sentence about the internship, the kind of problems you enjoy, and what you are deliberately building towards over the next few years.
AI applications, ML models and data tools — fast prototypes first, then properly.
Data, markets and messy situations — looking for the signal and the “so what”.
Notes, experiments and reflections. This website is part of that habit.
Placeholder. A retrieval-augmented assistant that answers questions from your own course notes — with citations, so you can check its work.
Placeholder. Forecasting weekly demand across product lines, then translating the model into a reorder recommendation a store manager could act on.
Placeholder. Should a mid-sized fintech enter a new market? A structured case: sizing, competitor mapping, risks and a clear recommendation.
Placeholder. Predicting which customers are likely to leave — and, more usefully, explaining why, so a team knows what to do about it.
Placeholder. Worked inside a client-facing data team — cleaning data, building analyses and turning findings into something decision-makers could use.
Placeholder. Supported a faculty-led project on applied machine learning — literature reviews, experiments and a lot of careful reading.
Placeholder. Ran workshops, organised a hackathon and learned that coordinating people is harder — and more interesting — than coordinating code.
How will AI change the way businesses make decisions — not just how fast they make them?
Exploring through: ML coursework, reading on decision science.
What makes a technology company defensible when the technology itself is cheap to copy?
Exploring through: case studies, business history.
When is a simple spreadsheet the better answer than a machine-learning model?
Exploring through: my internship, and a few over-engineered projects.
How do good consultants structure a problem nobody has framed yet?
Exploring through: case practice, frameworks, conversations.
learning to build
Started computer science: programming, data structures, first real projects.
learning to measure
Data analysis, machine learning and an internship inside a real team.
learning to decide
Business analysis, consulting thinking and clearer communication.
still writing this part
A role where technology and business decisions sit at the same table.