Manav Malhotra

I build products where machine intelligence meets physical constraints.

I'm interested in ML systems that understand consequences, especially where models have to operate in the physical world rather than just describe it. I co-founded a fintech company building a subledger (YC S20), built pricing systems in healthcare, and patented solar cell technology at Caltech.

Browse the writing archive

Cui Bono

Independent Project · Builder · Current

Cui Bono is a transparency tool for journalists, researchers, and citizens investigating who benefits from legislation. It uses LLMs to ingest bills, track revisions, analyze clauses, and connect provisions to industries and campaign donors. Most of the work is eval: making sure what the model says about a provision actually matches what the provision does.

LLMs, NLP

Financial Subledger

Blue Onion · Co-Founder · 2020–2026

I co-founded Blue Onion (YC S20), where we built a product that turned fragmented financial data into audit-ready records for finance teams. I built the data and reconciliation layer and defined the product requirements and measures. Built from 0-1 and scaled to hundreds of ecommerce customers.

Anomaly Detection, Supervised Learning, Best Fit Algorithms

Drug Pricing at Scale

SingleCare · VP of Data Science · 2019–2020

SingleCare provides free prescription drug benefits to everyone. I designed and built the pricing algorithm that powered it, pricing tens of thousands of drugs under generic-substitution and pharmacy contract constraints. Every price reached a patient filling a prescription, so the product had to balance contractual realities, competitive dynamics, and medication access, and keep working as the market moved.

Constrained Optimization, Reinforcement Learning, Supervised Learning

Marketplace Liquidity

Kinnek · Chief Data Scientist · 2014–2017

I was an early employee at Kinnek, a B2B marketplace where I built supervised learning models to predict qualified buyer leads and unsupervised methods for marketplace matching, focusing on the quality of marketplace liquidity. Backed by Thrive and Matrix.

Supervised Learning, Recommender Systems

Happy to compare notes.

The creation of a thousand forests is in one acorn. — Ralph Waldo Emerson