Together AI
Data AI · Open-source model infra
- HQ
- San Francisco, US
- Website
- together.ai
Together AI currently has 25 active AI-related job listings. The majority of these roles, 80%, are focused on serving infrastructure. Engineering is the dominant function with 21 listings, and the United States is the primary hiring country with 19 roles. Frequent tech tags include model serving, inference infrastructure, and fine-tuning. Over the last 30 days, Together AI posted 4 new AI roles, representing a 33% decrease compared to the previous 30-day period.
Currently tracking 22 active AI roles, down 17% versus the prior 4 weeks. Primary focus: Serve · Engineering. Salary range $121k–$300k (avg $227k).
Data AI · Open-source model infra
Together AI currently has 26 active AI-related roles in our index. The most common open titles are: Solutions Architect (2), AI Infrastructure Engineer, AI Researcher, Core ML (Turbo), AI infrastructure Engineer (SRE) Bangalore , Customer Support Engineer (Inference), India. Most positions are in Engineering and Research.
Together AI's active AI hiring is concentrated in: serving infrastructure (81%), post-training (12%), application (4%). These categories follow a seven-stage AI lifecycle: data, pre-training, post-training, serving infrastructure, agents, evaluation, and application.
Together AI is hiring AI talent in: United States (20 roles), Netherlands (3 roles), United Kingdom (1 role).
Job postings at Together AI most frequently mention: Production ML Systems, Inference Infrastructure, GPU Computing, LLM Inference, Text-to-Speech.
In the past 30 days, Together AI has posted 5 new AI-related roles.
| Title | Stage | AI score |
|---|---|---|
| Research Engineer, Core ML Research Engineer role focused on improving inference efficiency and unifying it with RL/post-training systems for production-grade AI APIs. The role involves end-to-end ownership of critical systems, translating frontier ideas into robust infrastructure, and shipping measurable improvements in latency, throughput, cost, and model quality at scale. | ServePost-train | 10 |
| Research Engineer, Large-Scale Training Research Engineer focused on scaling and optimizing large-scale training infrastructure for foundation models, integrating new architectures, and productionizing novel training methods. | Post-trainData |
| 9 |
| Research Intern RL & Post-Training Systems, Turbo (Fall 2026) Research intern focused on making post-training and reinforcement learning for large language models efficient, scalable, and reliable, by co-designing algorithms and systems at the intersection of RL, inference, and large-scale experimentation. | Post-trainServe | 9 |
| Research Intern, Model Shaping (Fall 2026) Research intern role focused on advanced post-training methods (supervised learning, preference optimization, RL), efficient neural network training techniques, and robust evaluation of foundation models. The team tailors open foundation models for downstream applications and develops new methods for efficient training and evaluation. | Post-trainPretrain | 9 |
| Research Intern, Inference (Fall 2026) Research intern focused on building efficient, scalable, and reliable serving systems for large foundation models, involving distributed inference, compiler optimization, and hardware optimization. | Serve | 9 |
| Frontier Agents Intern (Fall 2026) Research intern focused on building, aligning, and scaling frontier AI agent systems for complex multi-step tasks across text and speech. Projects involve developing new training methods, curating datasets, studying failure modes, and building scalable infrastructure for agent operations. | AgentPost-train | 9 |
| Research Engineer, Frontier Speculative Decoding Research Engineer focused on translating internal model training research into production-ready deployments by fine-tuning general-purpose models into specialized tools. This involves designing novel speculative algorithms, data curation, hyperparameter tuning, and checkpoint evaluation, with a focus on accuracy-efficiency tradeoffs for generative AI models. | Post-trainServe | 9 |