SG
LinkedIn

Saumya Gandhi

Member of Technical Staff

Bellevue, Washington, United States · 1 connection

Anthropic

Not on Unlinked · Imported profile. Is this you? Claim it

About

I am currently building OpenPipe: The End-to-End LLM Fine-tuning Platform for Developers. Prior to this, I completed my Masters in AI and Innovation at CMU. I am highly interested in creating impactful products that either use AI or enable others to create products that use AI. I've developed an AI guided journaling tool that was used by 500 students across CMU. I've contributed towards improving digital mental health technologies through work with Wysa and OptionsMD. I also have significant research experience with published NLP research at EMNLP and ACL, over my work with labs across IIIT Delhi (MIDAS Lab) and CMU (NeuLab, Dr. Graham Neubig). Ultimately, my goal is to found an AI-based startup focused towards positive societal impact. I am highly interested in discussing problems and solutions with this goal in mind. Currently, I am entirely focused on building OpenPipe! Want to: - Beat SOTA LLM completion performance? - Increase LLM reliability and reduce hallucinations? - Slash LLM costs up to 96%? - Lower LLM inference latency up to 300%? Our Platform simplifies and accelerates the entire fine-tuning lifecycle Use OpenPipe to: - Collect potential fine-tune training data at sca

Experience

Member of Technical StaffAnthropic
Sep 2025
Head of Model QualityOpenPipe
Jun 2024 to Sep 2025
Research AssistantOracle
May 2023 to Aug 2023
AI InternWysa
May 2022 to Aug 2022
Summer AnalystGoldman Sachs
May 2021 to Jul 2021
Explorer FellowMajor League Hacking
Feb 2021 to Apr 2021
Summer Research InternMIDAS: Multimodal Digital Media Analysis Lab
May 2020 to Jul 2020

Education

Carnegie Mellon University
Master of Science - MS, Artificial Intelligence and Innovation, 4.15
2022 to 2024
Visvesvaraya National Institute of Technology
Bachelor's degree, Computer Science and Engineering, 9.05/10
2018 to 2022
Pace Junior Science College
High School, Science, 89%
2016 to 2018