Together AI jobs, judged for you, free

What they do

Together AI is a cloud for AI engineers. It offers inference for open models, fine-tuning, reinforcement learning, pre-training, and GPU clusters. It serves more than 400 trillion tokens a month on a very large GPU fleet. The company started in 2022, and its team worked on FlashAttention and RedPajama.

What they value

• Open and responsible development.
• Empower innovation and do more with less.
• A strong sense of ownership and pragmatism.
• Curiosity about models.
• Comfort with ambiguity and frequent context switches.
• A "we ship it" attitude: engineers operate what they build.

Where they hire

Most roles are in San Francisco, with smaller groups in Amsterdam and London. Some Amsterdam roles are hybrid, two office days a week. India roles are remote in India or in Bangalore or Pune, in infrastructure engineering and technical support. Some support roles work weekend shifts.

Teams and roles

Engineering is the largest function and centers on GPU infrastructure. Languages are Go, Python, and Rust.

• Infrastructure: fleet automation for GPU clusters, Kubernetes controllers, Temporal workflows, storage such as Vast, Weka, and Ceph, and InfiniBand and RoCE networking.
• India engineering: an AI infrastructure systems engineer and a staff distributed storage engineer in Bangalore. Remote roles on the inference control plane and infrastructure agents.
• Infrastructure agents: AI agents that diagnose hardware failures, with knowledge graphs and retrieval.
• Product engineering: Playground, fine-tuning, batch, and evals in React and TypeScript, with Next.js and Tailwind as a plus.
• ML and research: voice model serving with TRT-LLM and SGLang, and large-scale training in Python and PyTorch on multiple GPUs.

Support engineers in Pune or Bangalore help inference and GPU cluster customers.

What they look for

Postings want an automation-first mindset and ownership of software in production. Research roles need hands-on multi-GPU training. Product managers for model APIs read code and build prototypes with AI tools. The careers page does not describe the interview steps.