Machine Learning Research Engineer
Flagship Pioneering, Inc. · Cambridge, MA
$120,000–$192,500from the description
Jul 27, 2026
Cambridge, MA
Jul 28, 2026
What this job asks for AI summary
This role at an early-stage AI startup focuses on researching and building machine learning systems — particularly LLM-based applications — aimed at improving psychological wellbeing. The engineer will design and deploy production-grade LLM pipelines, work on reinforcement learning from human feedback and agentic AI, and collaborate with interdisciplinary teams including psychologists. It suits someone with hands-on ML research and engineering experience who is comfortable in a fast-moving startup environment.
Mid level · Full-time
We read this from the posting text with AI. Skim the description below before ruling yourself out.
How this req sits in the market our data
Rare in this occupation — lead with these, and say what you built with them.
Most people in this occupation already list these. Still required — just not what gets you shortlisted.
What the occupation pays Median $122,874 (middle half $87,544–$162,374). This posting is about at that midpoint.
Estimated from BLS employment for this occupation and area, per-skill prevalence across our listing corpus, and published wage benchmarks — as of Jul 28, 2026. It is a model, not a headcount.
Why we read it this way (8)
Location is not specified in the posting; no CBSA or state could be determined. The role may be remote or based near Flagship Pioneering's Cambridge, MA headquarters, but this is not stated.
The SOC classification is a close call: the role blends ML research (15-2051 Data Scientists) with production software engineering of LLM systems (15-1252 Software Developers). The research and modeling emphasis — RLHF, fine-tuning, NLP, publications at ML venues — tips it toward 15-2051.
A Master's degree is listed as 'preferred' and explicitly noted as relaxable for exceptional candidates, so it is not treated as a hard requirement.
PyTorch, vLLM, and LangGraph are listed together under 'Proven experience deploying LLM-powered applications in production environments, including expertise in frameworks like…' — the 'like' qualifier signals these are named examples of a required capability rather than individually gated tools, but they are the only named frameworks and are treated as required given the firm 'proven experience' framing.
Reinforcement learning from human feedback and supervised fine-tuning appear in the role framing/intro rather than a formal requirements section, so they are marked as preferred.
Publications at top ML/NLP venues (NeurIPS, ICML, ACL, EMNLP) are explicitly called 'a bonus' and prior startup experience 'a plus' — both omitted from the skills list as they are not concrete technologies.
Seniority is assessed as Mid: no years of experience are stated, the degree requirement is relaxable, and while the work is technically sophisticated, the posting does not specify senior-level scope or leadership responsibilities.
Ignored 1 non-technology phrase(s) as skills (responsibilities/concepts, not named tools): reinforcement learning from human feedback.
Read the full posting
The employer publishes the full description on their own site — read it there ↗. Or sign in to read it here — it's free, and it also lets you track this application.