Research Scientist – Optimization

Percepta • Boston, MA

4d

About The Position

As a Research Scientist - Optimization at Percepta, you'll work at the intersection of AI research and real-world impact. Once a world model has compressed the operation and forecasters and user models have scored what happens next, someone has to actually choose the best action, under real constraints, at production scale. That's this role: combining modern machine learning with rigorous optimization research — test-time search, reinforcement learning, and classical OR — to close the loop from prediction to decision. Our optimizers are already running in production, allocating scarce resources against real operational constraints and generating measurable multi-million dollar impact. We've also built internal benchmarks with real, undisputed ground truth — replayable operations where a decision system's performance can be scored — to rigorously validate new methods before they ever touch a live customer.

Requirements

• PhD in Computer Science, Operations Research, Industrial Engineering, or Applied Mathematics or have equivalent research/industry experience.
• Have depth in operations or mathematical optimization (LP/MIP/MINLP, CP, stochastic/robust optimization).
• Have experience in novel machine learning techniques for Operations Research, including test-time search and reinforcement learning for sequential decision-making.
• Are comfortable implementing fast real-time optimization systems, debugging large-scale optimization systems, and/or designing benchmarks with real, defensible ground truth.
• Are motivated by impact in critical industries including healthcare, supply chains, energy, and finance.
• Have a proven track record of execution.
• Are an excellent communicator with both technical and non-technical stakeholders.
• Enjoy extreme ownership.
• Are passionate about AI's transformative potential.

Responsibilities

• Set and drive ambitious research programs that expand what's achievable in data-driven decision-making, building on the forecasts and user models produced by our research tracks.
• Invent new optimization methods for high-impact problems such as planning, scheduling, routing, pricing, and inventory — combining test-time search and reinforcement learning with classical LP/MIP/CP formulations.
• Build high-fidelity simulators and rigorous, replayable benchmarks (in the spirit of the subway challenge) that mirror real-world constraints, uncertainty, and multi-objective trade-offs, and that let us validate a decision before it touches a real patient, member, or asset.
• Push optimizers from benchmark wins into production systems that hold up against real operational stakes — not just a demo, but something that runs fast in real-time and re-learns from outcomes every week.
• Bridge research into practice by partnering with our engineers to rapidly prototype solutions and implement successful research ideas across live customer engagements.

What This Job Offers

• Job Type: Full-time
• Career Level: Senior
• Education Level: Ph.D. or professional degree