Ellysha Fatima
Open to AI engineering roles

I build AI systems
that people actually use.

Machine learning is the easy half. The rest is the interface, the failure modes, and the rules an organisation needs before it can put any of it in front of real people. That's the half I work on.

Storybook Studio
page 1 of 6

Running live, not a screenshot.

Selected work

All live · try them yourself

01

Storybook Studio

LLM · Image generation

Writes an illustrated children's book pitched to a specific reading level, with vocabulary and activities to match.

The hard part is character consistency — keeping the same dragon recognisable across six generated pages.

Open it→
02

AI Governance Sandbox

Policy · Risk modelling

Classifies an AI system against the EU AI Act and returns the controls, sign-offs and review cadence that follow.

The classification is deterministic code, never model output. A compliance tool that answers differently twice is worse than none.

Open it→
03

Sensor-Free Solar Tracker

Edge ML · Simulation

Derives panel angles from astronomy alone — no light sensors — then distils the physics into a 1.7 KB neural net.

The network trains live in your browser and recovers the analytic model to within about a degree.

Open it→