Research Engineer

Sourced from simplify.jobs

SuperAnnotate AI is hiring a Research Engineer to drive frontier-facing research, from reimplementing papers and benchmarks to building reproducible evaluation pipelines. You'll own projects end to end, including RL/agentic environment builds and novel multimodal benchmarks, turning research directions into customer datasets, pilots, and publications. The role is full-time and hybrid in San Francisco.

• Independently identify and reimplement relevant papers, benchmarks, and methods
• Build and own processes to reproduce prior work and improve on it
• Own projects end to end, including scoping, MVP implementation, and validation
• Translate ambiguous requirements into concrete, testable research plans
• Validate ideas through hands-on implementation, annotation, and evaluation
• Turn research into tangible outputs such as datasets, pilots, or publications
• Assess technical feasibility of new opportunities and shape proposals

• MS or PhD in ML, CS, or a related quantitative field, or equivalent research experience
• Real ML depth: understanding of model training and evaluation, not just API usage
• Hands-on experience with RL/agentic systems, AI/ML evaluation and benchmarking, or multimodal ML
• Strong Python and ability to build and ship experiments independently
• High autonomy and clear technical writing

• Publication track record (first-author preferred)
• Experience with agent or multimodal benchmarks such as OSWorld, MMMU, WebArena, or SWE-bench
• Familiarity with reward modeling, reward hacking, or verifier/judge reliability
• Familiarity with synthetic data generation or human-in-the-loop workflows
• Experience with cloud infrastructure and containerized environments
• Deep RL background