// About

A curious mind
learning to make
machines think.

I'm Haarish Jayakumar — a learning AI engineer who spends most days deep in notebooks, papers and agent traces, chasing the frontier one experiment at a time.

My focus is core machine learning, large language models and agentic AI. I love the moment when a messy problem clicks into a clean gradient — and the moment when an agent finally does the thing.

Beyond building for myself, I also offer AI engineering services — helping teams and founders prototype, ship and evaluate intelligent systems that feel less like magic and more like craft.

Core Machine Learning

Building intuition from the math up — linear models, deep networks, optimization, evaluation.

Large Language Models

Prompting, RAG, fine-tuning and evals. Turning transformers into useful products.

Agentic Systems

Tool-using agents that plan, reason and act — with memory that actually holds.

Shipping Real Things

Prototypes that run, APIs that scale, interfaces humans actually enjoy.

"Learn in public. Build in the open. Let the models do the talking."