Singapore · data engineering · AI systems · software engineer

Prittam Ravi

Production data systems and AI infrastructure, building for both.

I work at a hedge fund in Singapore on data pipelines that have to stay up. Outside that I look at AI infra and new products in the space: small scripts, toy systems, papers, and whatever tool or model release is worth trying that week.

I started reading more papers when I noticed I could agree with an argument in a meeting and still struggle to explain it an hour later. Writing is the fix: I can't hand-wave past the parts I don't get. Most of what's on this site started that way.

Lately that's meant context and memory for agents, evals that survive contact with a real codebase, harnesses I'm still trying to understand properly, decent tracing when the model does something weird, and agents that don't fall apart after the tenth step.

Building at work

data pipelines production tooling Python / SQL

Exploring

ai-infra memory systems observability agents harnesses

Writing about

papers systems distributed systems

Outside work

Sports

Learning to box — still getting hit more than landing punches. Dodgeball on weekends. I like sports that punish slow thinking, and I'm always open to trying something new.

Adventures

Hiking, skiing, white-water rafting. I travel when I can and lean toward trips where something can go wrong but usually doesn't. I want to see and understand as much of the world as I can.

Reading

Papers, technical essays, and whatever's been sitting in my reading list longest. Mostly things that change how I see a problem I'm already thinking about.

Writing AI tooling, databases, and systems notes Projects data pipelines, CLIs, aggregation, and knowledge tools Experience financial datasets, workflow automation, and product tooling