Member of Technical Staff, Applied Research at abundant, Bengaluru
Open. Posted .
Applied reinforcement-learning research builds agentic tasks, verifiers and evaluation systems, then turns research into production-scale data pipelines for frontier AI models.
Applying opens the employer's page. Want jobs like this sent to you each morning? Open Elip.
| Location | Bengaluru, Karnataka |
|---|---|
| Work mode | On-site |
| Experience | 5–8 years |
| Pay | Not stated by the employer |
| Role | AI Engineer |
| Skills asked for | Python, Reinforcement Learning, Deep Learning, Post-training, Distillation, Agent harnesses, Evaluation, Distributed execution |
| Posted | 20 Nov 2026 |
Elip's summary
Works as an applied research engineer at a lab scaling reinforcement learning for safe, reliable agentic capabilities. Builds agentic tasks, verifiers, evaluation systems, and production research pipelines, applying reinforcement learning, deep learning, and Python. Ships post-training research and develops data strategies for model alignment and safety.
The job, in the employer's words
ABOUT ABUNDANT
Abundant is an applied research lab focused on scaling reinforcement learning for safe and reliable agentic capabilities. We are an extremely talent-dense team of researchers, roboticists, founders, and operators whose work includes the Waymo Driver.
THE ROLE
As a Member of Technical Staff on Applied Research, you'll read the latest papers, decide which ideas are worth betting on, quickly reproduce them, and then build the pipelines that run them at production scale. You'll own the whole path from an arXiv link, to a system that produces data every day, to delivery to our customers at the frontier of artificial intelligence.
WHAT YOU’LL DO
- Stay current with the latest research and techniques. Read widely across RL, post-training, agents, evals, and synthetic data, and decide what's worth building.
- Design tasks and verifiers that stay hard. Build long-horizon agentic tasks that are still difficult a model generation from now, with graders where partial credit means something and the obvious exploit does not score.
- Diagnose failure modes. Read trajectories until you can say why a model fails a class of work, then turn that into the data or reward signal that fixes it. This is the core loop of the job.
- Model alignment and data strategy. Partner with the world’s most advanced AI research teams on datasets and benchmarks that shape how frontier models behave, focused on alignment, safety, and reward signal design.
- Autonomous problem selection. Identify, scope, and run long projects, choosing the problems that matter most for scaling data toward AGI.
- System infrastructure. Work with engineering on data pipelines, internal tooling, and deep learning implementations. If a workflow produces bad data, you fix the workflow.
WHO YOU ARE
- You write a lot of code. Strong software engineering and deep Python, comfortable owning a system end to end and debugging through layers you did not write.
- You have built evaluations, RL environments, agent harnesses, verifiers, or execution sandboxes, and you have opinions about which of them can be trusted.
- You have shipped research into production, particularly post-training, distillation, and high-stakes evaluation.
- You have worked on large-scale agent systems: orchestration frameworks, tool APIs, distributed execution, observability, and logging.
- You get to ground truth in an unfamiliar domain by building something and reading outputs, not by reading about it.
- You notice the misspecified task, the ambiguous prompt, and the grader that can be gamed.
- You have a thoughtful perspective on the societal and safety impacts of deploying general-purpose AI systems.
NICE TO HAVE
- A PhD (CS, ML, NLP) and publications at NeurIPS, ICML, or ICLR. Neither is required.
- Work on public benchmarks or task suites, for example Terminal-Bench, SWE-bench-style harnesses, or HCAST.
- RLHF or RLAIF experience and a view on how training choices show up in model behavior.
- Experience running work through large expert or annotation operations.
About abundant
Headquartered in San Francisco, founded in 2024.
- People in Elip's index who work here
- 31
- Typical time at abundant
- 9 months
- Typical career length so far
- 4.0 years
Totals from public professional profiles in Elip's index; any group under ten people is left out.
Similar open jobs
About this page
Elip lists jobs that employers publish on their own careers sites, never copies from job boards. This one comes from abundant's listing on Ashby, posted 20 Nov 2026. The summary is Elip's; the description is the employer's. When the job closes, this page says so.