Functional Focus: Data Engineering, Real-Time Streaming, and HPC Job Orchestration
- Scientific Data Architecture: Design crash-safe, binary data formats capable of handling high-speed incremental writes from compute jobs and concurrent reads for query-time calculations.
- Operational Data Modeling: Design and maintain a relational database schema via an async ORM, backend-agnostic across database engines.
- HPC Workload Management: Manage the lifecycle of batch and interactivecompute jobs, handling subprocess monitoring, status polling, and cluster filesystem coherency.
- Real-Time API & Streaming: Build async backend services and long-lived, server-pushed event channels with backpressure handling, and enforce role-based access control on all APIs.
- AI Service Integration: Integrate and orchestrate calls to an AI/analytics service from the backend, coordinating with application state.
Requirements
Mandatory:
- Python 3, async/await
- FastAPI (or similar async Python web framework)
- Relational DB modeling, async ORM (e.g. Tortoise, SQLAlchemy), SQL
- Server-Sent Events or WebSockets, backpressure handling
- Role-based access control (RBAC), token-based auth (JWT/OAuth)
- Subprocess management, batch job lifecycle/status polling
- Binary/streaming file format design, crash-safe writes
- REST/SDK integration with an external AI/LLM service
Optional:
- Go or another async-capable backend language
- GraphQL
- PostgreSQL administration, Alembic/migrations tooling
- Message queues/brokers (Redis, RabbitMQ, Kafka)
- Air-gapped/offline deployment experience
- HPC schedulers (SLURM, PBS), Linux cluster filesystems
- NumPy/columnar formats (Parquet, HDF5)
- Tool-calling / function-calling orchestration patterns
Benefits
We offer great career growth, ESOPs, Gratuity, PF and Health Insurance.