Research Engineering Manager, STEM
Listed by Jobgether for an employer it does not name
- LocationBrazil
- Work typeRemote
- EmploymentFull-time
- Posted18 November 2026
- SourceListed by Jobgether (Lever)
Last checked1 h ago: still listed on Jobgether. First seen 5 h ago.
Moves fastWhen Jobgether takes a role down early, it is usually about 8 days after posting. Apply soon.
At a glance
- Experience8+ years
- EducationMaster's degree
- Skills and toolsMachine learning, LLMs, Python, PyTorch, Spark, Kubernetes, AWS, Google Cloud
Read from the ad by RoleFinder, with rules and an AI reading; the description is the source.
About the role
This position is listed on behalf of a partner company, who manages all applications and next steps. Our partner is looking for a Research Engineering Manager, STEM based in Brazil.
As a Research Engineering Manager, STEM, you will lead a team of research engineers developing advanced datasets, synthetic data systems, and evaluation benchmarks for next-generation AI models. Working at the intersection of machine learning research and engineering, you will collaborate closely with research scientists and leading AI laboratories to improve model capabilities in coding and advanced STEM reasoning. You will oversee the design and implementation of scalable data generation pipelines, rigorous evaluation methodologies, and reinforcement learning environments. Alongside managing and developing your team, you will remain deeply involved in technical execution, contributing directly to complex engineering challenges and research decisions. You will also help define research priorities, establish quality standards, and ensure that experimental results are reliable, reproducible, and meaningful. This is an opportunity to influence the direction of frontier AI research while delivering measurable improvements to advanced AI systems.
Accountabilities:
Lead the research engineering function responsible for developing high-quality synthetic datasets, reinforcement learning environments, and advanced evaluation benchmarks. You will combine technical leadership, people management, and hands-on research engineering to ensure that projects deliver scientifically rigorous results and measurable improvements in model performance.
-
Build, manage, and develop a high-performing team of research engineers, defining hiring profiles, leading recruitment, and selecting professionals with strong research judgment and engineering expertise.
-
Establish the team's operating structure, including project prioritization, planning, experiment reviews, technical documentation, and clear ownership of deliverables.
-
Coach engineers on experimental design, software quality, research methodology, and technical communication while providing regular feedback and conducting performance evaluations.
-
Define and maintain engineering standards for evaluation methodologies, data quality, and reproducibility, ensuring that datasets and experiments can be regenerated from documented code, configurations, and random seeds.
-
Partner with research scientists to translate research hypotheses into structured experiments, including baseline selection, ablation studies, evaluation criteria, and measurable success outcomes.
-
Contribute independent research ideas and help establish the team's technical and scientific priorities.
-
Translate research prototypes into scalable production systems while preserving their original scientific objectives.
-
Design and oversee synthetic data pipelines covering generation, verification, filtering, deduplication, and versioning through programmatic generators, LLM-based generation, automated verification, and expert review.
-
Evaluate synthetic data quality through fine-tuning and reinforcement learning experiments, measuring how factors such as difficulty, diversity, correctness, and coverage influence model performance.
-
Develop robust quality assurance systems using gold datasets, LLM-judge calibration, expert annotations, sampling audits, and agreement metrics.
-
Identify capability gaps in existing AI models and develop new benchmarks and evaluation methods to measure areas not adequately covered by current assessments.
-
Own benchmark methodology end to end, including task taxonomies, difficulty calibration, grading systems, contamination controls, held-out datasets, and statistical confidence intervals.
-
Build and maintain evaluation harnesses, including agent-based and sandboxed execution evaluations, ensuring deterministic, reproducible, and trustworthy results.
-
Assess experimental findings and determine whether observed model improvements represent statistically meaningful gains or measurement noise.
-
Collaborate with research scientists to publish datasets, benchmarks, and research findings at leading conferences such as NeurIPS, ICML, and ICLR.
-
Remain hands-on with complex technical work, dedicating at least one-third of your time to coding, prototyping, debugging, and improving critical research infrastructure.
-
Serve as the technical reference for complex methodology and engineering challenges, including verifier disagreements, evaluation inconsistencies, and data quality issues.
-
Act as a technical counterpart to researchers at leading AI laboratories, defining project requirements, acceptance criteria, and communicating results and technical trade-offs.
-
Coordinate with strategic project leads, delivery teams, and engineering partners to translate research advancements into client solutions and scalable platforms.
Requirements:
You are an accomplished research engineering leader with extensive experience in machine learning, synthetic data generation, and AI evaluation systems. You combine strong technical expertise with proven people management capabilities and a rigorous scientific mindset. You are comfortable leading research teams while remaining actively involved in engineering implementation, experimental design, and technical decision-making.
-
At least 8 years of hands-on experience in machine learning or research engineering, including at least 5 years in industry.
-
At least 3 years of direct management experience leading research engineers, machine learning engineers, or research teams, including hiring and performance management.
-
Proven experience building and managing synthetic or curated data pipelines in production, with measurable impact on model training and performance.
-
Strong understanding of synthetic data generation, verification, filtering, and quality assessment methodologies.
-
Demonstrated experience designing and delivering LLM evaluations or benchmarks used by other teams or organizations to make model development decisions.
-
Deep understanding of evaluation challenges, including data contamination, benchmark saturation, judge bias, inter-annotator agreement, statistical variance, and confidence intervals.
-
Strong experimental design skills, including the ability to define meaningful baselines, plan ablation studies, and distinguish genuine performance improvements from statistical noise.
-
Advanced Python programming skills and production experience with PyTorch or JAX.
-
Working knowledge of modern AI post-training methodologies, including Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), Proximal Policy Optimization (PPO), Group Relative Policy Optimization (GRPO), and reinforcement learning with verifiable rewards.
-
Experience with distributed data processing and large-scale evaluation or training workloads using technologies such as Ray, Spark, or Kubernetes.
-
Familiarity with cloud infrastructure, particularly AWS or GCP.
-
Ability to establish technical direction, define research priorities, and guide a team beyond simply executing an existing roadmap.
-
Excellent written and verbal communication skills, with the ability to present complex technical findings to research scientists, engineering teams, and non-specialist stakeholders.
-
Daily experience using AI-powered coding tools such as Claude Code, Cursor, or GitHub Copilot, with the ability to critically evaluate and validate AI-generated outputs.
-
Master's or PhD degree in Artificial Intelligence, Machine Learning, Computer Science, Statistics, or a closely related discipline. A PhD is preferred, although exceptional equivalent research experience will also be considered.
Strongly Preferred:
-
Previous experience at a frontier AI laboratory or leading industry research organization, particularly in post-training data, model evaluation, or reward modeling.
-
Experience contributing to or developing publicly recognized benchmarks or datasets used by the research community.
Nice to Have:
-
First-author research publications at leading conferences or journals, including NeurIPS, ICML, ICLR, ACL, or EMNLP.
-
Experience designing reinforcement learning environments and verifiers for agentic tasks, including coding agents, tool use, computer use, and long-horizon workflows.
-
Knowledge of data valuation and attribution methodologies to measure the contribution of individual data segments to model performance.
-
Advanced expertise in a STEM discipline such as mathematics, physics, chemistry, or biology.
-
Experience managing technical leads or other managers, or coordinating multiple concurrent research programs.
-
Experience directing large expert networks through detailed specifications, evaluation rubrics, and rapid feedback processes.
Benefits:
-
Opportunity to work directly with leading AI laboratories and enterprises on advanced model development, post-training, evaluation, and synthetic data initiatives.
-
Direct involvement in building high-impact datasets, reinforcement learning environments, and innovative benchmarks for next-generation AI systems.
-
Opportunity to influence how frontier AI models are evaluated and how progress in coding and STEM reasoning is measured.
-
High level of autonomy and ownership in a fast-moving, research-driven environment.
-
Collaboration with experienced AI researchers and technology professionals from leading organizations such as Google, Meta, and Amazon.
-
Opportunities to contribute to cutting-edge research and publish findings at prestigious conferences, including ICLR, ICML, and NeurIPS.
-
Exposure to enterprise AI applications, translating frontier research innovations into practical business solutions.
-
Opportunity to work at the forefront of AI innovation in a dynamic, startup-paced environment.
-
Inclusive and diverse working culture that values professional growth, authenticity, and collaboration.
How Jobgether works:
We use an AI-powered matching process to ensure your application is reviewed quickly, objectively, and fairly against the role's core requirements. Our system identifies the top-fitting candidates, and this shortlist is then shared directly with the hiring company. The final decision and next steps (interviews, assessments) are managed by their internal team.
We appreciate your interest and wish you the best!
Why Apply Through Jobgether?
Data Privacy Notice: By submitting your application, you acknowledge that Jobgether will process your personal data to evaluate your candidacy and share relevant information with the hiring employer. This processing is based on legitimate interest and pre-contractual measures under applicable data protection laws (including GDPR). You may exercise your rights (access, rectification, erasure, objection) at any time.
#LI-CL1
Get open roles ranked against your CV
Add your CV once. RoleFinder ranks the open roles where you want to work, plus remote ones, against it, shows your best 12 with a one-line reason each, and emails you new matches once a week. The file is read once by AI and not kept; employers never see it.
Sign in with Google or email to start. How we handle your CV