Aastha Jain

Aastha Jain

Long curiosity, short certainty.

scroll

What I do

I trade options. I started out building the tools and data systems around the desk, and slowly found myself spending more time on research, risk, and the actual trade decisions.

Before trading, I spent a lot of time around research. At Columbia, I worked on multi-objective LLM alignment; at Microsoft Research, on adaptive retrieval for RAG systems; and at TIFR, on deep learning methods for rare event simulation. I left Columbia's PhD program in operations research with a master's degree, and have a B.Tech from IIT Delhi.

I also started Weaverly, a project for helping women find real friendships in their city.

Recent posts

February 2026
OCaml 5, Effect Handlers, and the Slow Death of Colored Functions
A gentle introduction to effect handlers, why async changes spread through a call stack, and the alternative OCaml 5 offers.
December 2025
How I Think About the Kelly Criterion
A useful answer to a narrow question: if you have an edge, how much risk should you take without overbetting it?
June 2025
A Day in the Life of a Neural Network
What if a neural network were human? A playful exploration of training, inference, and the daily struggles of gradient descent.
view all posts →

Selected publications

A Deep Learning Approach for Rare Event Simulation in Diffusion Processes
Hult, H., Jain, A., Juneja, S., Nyquist, P. and Vijayan, S.
Winter Simulation Conference (WSC), IEEE — 2024
Comparing skill of historical rainfall data based monsoon rainfall prediction in India with NCEP-NWP forecasts
Narula, A., Jain, A., Batra, J., Juneja, S.
arXiv:2402.07851 — 2024 · NeurIPS Workshop 2025
Heuristic-Based Service Allocation and Price Determination in Cloud Manufacturing Operations
Mishra, A., Jain, A., Alfas, M., Shriyam, S., and Kumar, A.
IEEE CASE — 2023
view all publications →