Apply custom AI criteria to text and JSONL data from local files, SSH, or HDFS. Combine classification, scoring, and yes/no judgments, review predictions, record human corrections, and automate follow-up with Python Actions. Run manual or scheduled batches, or monitor local and SSH files in realtime. HDFS supports batch reads. Use self-hosted Laya or the optional paid Jev API. Includes Docker Compose setup and sample datasets.
