Baseten Tech Stack
AI infrastructure platform for deploying and scaling large language models
Baseten operates an inference-focused AI infrastructure platform built on NVIDIA GPUs, Kubernetes, and custom performance optimization (vLLM, TensorRT, CUDA). The stack reveals a systems-level engineering focus: adoption of InfiniBand and RDMA signifies movement toward high-performance interconnect, while the project list spans frontier model APIs, billing platform integrations, and internal ML optimization — indicating product maturation beyond early-stage tooling. Hiring is engineering-heavy (55 of 102 roles) with senior-skewed seniority distribution, paired with active GTM hiring, suggesting scaling from technical adoption to enterprise sales motion.
Tech Stack 82 technologies
What Baseten Is Building
◆Challenges
- Capacity management
- Improving engineering productivity
- Scalable infrastructure
- Scaling large models
- Billing strategic lever
- Deploying ml models
- Building foundation and processes for customer-facing teams
- Multi-tenant isolation
- Enhancing ci pipeline speed
- Optimizing ml model inference performance
▲Active Projects
- Model apis for frontier models
- Baseten inference stack
- Playbook style deployments
- Go-to-market strategy for fastest growing customer segment
- Building foundation for customer-facing teams
- Compensation plan design and rollout
- The fastest, most accurate whisper transcription
- Customer pocs
- Billing platform and orb integrations
- Internal tooling for quoting and approvals
Hiring Activity
Department
Seniority
Notable leadership hires: Head of Legal Operations