Kinjal

Securely isolated, containerized environment for executing LLM-generated data science scripts.

Persistent relational database archiving session threads, metadata, and dataset schemas.

Core FastAPI server executing the LangGraph engine and orchestrating multi-agent workflows.

High-performance web server container hosting and serving the compiled React frontend application.

FastAPI service running PyTorch inference to process NMT and export attention tensors.

Multi-stage build serving a React dashboard via Nginx to visualize real-time PyTorch inference.