Pallav Mahapatra
Open to workSenior Data Scientist · Amazon · Seattle, WA
I've spent 9 years at Amazon working at the intersection of causal inference, experimentation, and generative AI. I build systems that help product teams make decisions they can actually trust. This site is where I document what I've learnt.
Background
I started my career building data pipelines — not glamorous, but it taught me what it means for data to be trustworthy. I spent five years as a Lead BI Engineer at Amazon, owning KPI reporting for a $10B platform. Watching teams make confident decisions from numbers I knew were noisy pushed me toward a harder question: when a metric moves, did something you did actually cause it?
That question led me into causal inference — and to a graduate program in Data Science while working full time. As a Data Scientist I've since built distributed causal platforms, run experiments at scale, and spent the last year designing how Amazon evaluates generative AI before it ships. The core problem is always the same: how do you measure something rigorously enough to actually act on it?
Experience
Sr. Data Scientist — Customer Experience
Amazon
Built end-to-end GenAI evaluation systems for shopping assistants. Designed LLM-as-Judge frameworks, multimodal pipelines, and large-scale A/B experiments measuring causal lift across 450K products.
Data Scientist — Seller Experience
Amazon
Built a distributed causal inference platform in PySpark to estimate treatment effects at GB scale. Applied DoubleML, DR Learners, synthetic control, and WGAN-based counterfactual generation.
Lead Business Intelligence Engineer
Amazon
Owned KPI reporting for a $10B lead management platform. Built metric layers supporting 50+ analysts, reduced query latency 13%, and led the team through experimentation readiness.
Education
MSc, Data Science
Liverpool John Moores University, UK
B.E., Electronics & Telecommunication
University of Mumbai, India
Open to opportunities
Looking for the right next problem to work on.
I'm interested in roles and collaborations where rigorous measurement actually shapes product decisions — teams that care about getting causality right, not just shipping dashboards. If that sounds like your team, I'd love to connect.
What I bring
9 years at Amazon. End-to-end DS work: from data pipelines to causal models to GenAI in production.
Where I thrive
Teams that run experiments, care about measurement quality, and want to understand why metrics move.
Also open to
Research collaborations, guest writing, and technical conversations with people building in this space.