a
adi_gupta2

Aditya Gupta

@adi_gupta2

AI evaluation

India
Engels, Hindi
Sommige informatie wordt in het Engels weergegeven.
Over mij
I am a B.Tech. candidate with hands-on experience creating and reviewing Terminal bench and SWE bench benchmark tasks for frontier AI agents and building Python ML/RAG systems. I specialize in AI agent evaluation, MLOps, and responsible AI safety guardrails using tools like FastAPI, MLflow, Docker, and AWS.... Lees meer

Skills

a
adi_gupta2
Aditya Gupta
offline • 
Gemiddelde reactietijd: 1 uur

Bekijk mijn diensten

Programmering en technologie
I will create genuine hard terminal bench and swe bench benchmark tasks

Werkervaring

AirDawg_Labs

AI Systems Evaluation Intern

AirDawg Labs • Parttime

Jul 2026 - Present • 4 mos

Created and submitted 32 hard Terminal-Bench 2.0 tasks through Snorkel for frontier AI agent training and evaluation. Selected as a reviewer for the Sentinal Ultra project, conducting quality and task-readiness reviews for SWE-bench benchmark tasks. Submitted 21 tasks for SWE-bench with 17 accepted and reviewed 83 submitted tasks, implementing necessary fixes.