Member of Technical Staff, Post-Training

Ricursive Intelligence • Palo Alto, CA

5d

About The Position

Ricursive Intelligence is a frontier AI lab developing self-improving systems, beginning with chip design. The company aims to connect AI development with the hardware that powers it and accelerate progress toward artificial superintelligence. It is backed by $335M from Sequoia, Lightspeed, DST, and NVIDIA Ventures.

The company is seeking post-training researchers and engineers. A five-person team trained Ripple-65B Preview from scratch and achieved a U.S. open-weight state-of-the-art score on agentic coding, using internally built synthetic data pipelines and substantially fewer rollouts than the prior state of the art.

In this role, you will own core post-training work spanning synthetic data, reinforcement-learning environments, and evaluation. Your work will directly affect the model as the company scales its efforts at the intersection of AI models and chip design.

Requirements

• Degree in computer science, electrical engineering, mathematics, or a related field, or equivalent practical experience.
• Hands-on post-training experience in SFT, RL, synthetic data, and environments at a frontier AI lab or equivalent.
• Deep fundamentals in transformer internals and distributed training.
• A shipped result, such as a post-training run, model release, or research outcome that moved a benchmark or product.

Nice To Haves

• Research track record in RL, NLP, or machine learning, such as publications at NeurIPS, ICML, ICLR, or COLM.
• Experience building RL environments for complex frontier coding and agentic benchmarks.
• Experience shipping synthetic data generation and filtering at scale.
• Experience shipping RL infrastructure, such as asynchronous rollouts and inference engines like vLLM or SGLang.
• Experience navigating ambiguity and making design decisions across the stack.
• Interest in applying LLMs to hardware and chip design; prior experience is not required.

Responsibilities

• Design and iterate on supervised fine-tuning (SFT) and reinforcement-learning (RL) recipes for large-scale models, focusing on compute efficiency and steep, stable RL improvement.
• Build and scale synthetic-data pipelines, including generation, quality filtering, and coverage metrics.
• Create long-horizon synthetic environments and task curricula for agentic coding and reasoning, adding harder tasks as model capability improves.
• Improve RL stability at scale through recipe and stability work.
• Build evaluations that track agentic capability, catch regressions, and inform recipe decisions.
• Own experiments end to end on large GPU clusters, including designing, launching, debugging, and analyzing runs.
• Partner with teammates to improve post-training efficiency on the company's training and serving platforms.

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level