Amir Haghighat is a co-founder and the chief technology officer of Baseten, which runs inference for open and custom AI models. Haghighat's posts on the Baseten blog include a 2025 write-up, co-authored with Tri Dao, Abu Qader, Bryce Dubayah and Philip Kiely, on optimizing GPT OSS 120B for latency and throughput on launch day.12
At a glance
- Current role
- Co-founder and Chief Technology Officer
- Company
- Baseten
- Sector
- Company
- Fields
- Inference and serving
- Work listed
- 1
- Sources cited
- 4
- Profile reviewed
Baseten
Haghighat is listed on Baseten's blog as CTO and co-founder. In March 2026, Haghighat announced that Sameer Paranjpye had joined Baseten to lead its engineering organization.13
Haghighat's technical writing for the company includes an August 2025 account of day-zero performance work on GPT OSS 120B, written with Tri Dao, Abu Qader, Bryce Dubayah and Philip Kiely, and a December 2025 post with Alex Ker on infrastructure lessons for running AI in production, covering reliability, speed and developer control.24
Work and contributions
GPT OSS 120B performance optimization on Baseten Proprietary product
First-listed author of Baseten's August 2025 write-up on running GPT OSS 120B at high throughput on NVIDIA GPUs, covering first inference, compatibility bug fixes and model configuration tuning.2
Timeline
Where they work
- Baseten 3 people in the wiki
Related people
- Pankaj GuptaAlso at Baseten
- Tuhin SrivastavaAlso at Baseten
- Yagil BurowskiInference and serving
- Luis Héctor ChávezInference and serving
- Matthew ZeilerInference and serving
- Horace HeInference and serving
- Hani JamjoomInference and serving
- Priya NagpurkarInference and serving
- Ian BuckInference and serving
- Woosuk KwonInference and serving
Suggest a correction or removal
Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.
References
4 sources · profile reviewed Oct 22, 2026 · data updated Oct 22, 2026