Research Engineer

The role

You will build and scale modeling capabilities by developing large-scale training and inference infrastructure and reliable distributed systems. You will work embedded in customer environments to deploy production systems that turn research ideas into real applications.

What you will do

• Prioritize infrastructure and tooling work to connect research vision to business value.
• Prototype, build, and deploy AI systems for real users and production constraints.
• Iterate based on user feedback and partner with stakeholders.
• Embed within customer environments as a technical partner on applied AI.
• Contribute new features, patterns, and abstractions to the Mosaic platform.
• Translate learnings into reusable playbooks with Embedded Product Managers.

Required qualifications

• Experience with high-performance, large-scale distributed systems.
• Ability to scale research ideas into production.
• Proven track record of execution in ambiguous, fast-moving environments.
• Excellent communicator with both technical and non-technical stakeholders.
• Enjoy extreme ownership.
• Passionate about AI's transformative potential.

Nice to have

• Large-scale model training and inference.
• Implementing LLM post-training recipes.
• Deeply familiar with vLLM/SGLang, Ray and Kubernetes.
• Distributed checkpointing and multi-node/multi-GPU training.
• Asynchronous training and inference — VeRL, ROLL, SkyRL, AReaL, or asynchronous RL libraries like CleanRL.

Benefits

• Competitive compensation
• Equity
• 401(k) matching
• Comprehensive health and family benefits
• Lunch every day

About the company

Percepta is a direct partnership with General Catalyst, a global transformation and investment company. The mission is to transform critical institutions with applied AI.