Fig 1
DataOnCallEnv
RL environment for debugging data pipelines
Python
FastAPI
Hugging Face Spaces
Reinforcement Learning Benchmark
System Highlights
- >Partial observability environment with simulated Airflow/dbt logs
- >FastAPI backend (100+ SQL executions/session)
- >Multi-metric reward system (4 signals)
- >Reduced unnecessary queries by 25–30%
- >Accepted by Meta Pytorch hackathon