Staff Engineer, API Platform

About the role

Sarvam is building India’s sovereign AI platform and needs a Staff Engineer to own its API Platform end‑to‑end. You will lead the redesign from Python/FastAPI to Go, ensure reliability, performance and scalability for millions of daily API calls.

Key responsibilities

• Design and evolve the platform architecture from edge request to response, handling HTTP and WebSocket traffic.
• Lead the Python‑to‑Go rewrite, set patterns and maintain reliability during migration.
• Engineer audio pipelines (TTS chunking, sample‑rate handling, VAD‑based silence detection) and vision pipelines (OCR, layout detection, VLM integration).
• Build streaming infrastructure with WebSockets, queue‑based batch processing, back‑pressure and sub‑second latency SLOs.
• Implement the commercial layer: metering, billing, prepaid wallet management and rate‑limiting.
• Ensure observability (logging, metrics, tracing) and develop integration‑testing harnesses for rapid, safe releases.

Required profile

• 7+ years building production backend systems at scale.
• Strong Go experience in production, designing and operating high‑load services.
• Comfortable with Python and FastAPI for the current codebase.
• Proven track record with high‑scale, low‑latency, multi‑tenant distributed systems.
• Hands‑on experience with real‑time streaming (WebSockets, long‑lived connections, back‑pressure).
• Deep knowledge of PostgreSQL and Redis (schema design, query performance, caching).
• Experience running services on Kubernetes in production.
• Senior‑level judgment on building vs. buying, optimizing vs. shipping, and mentoring peers.

Required skills

• Go
• Python
• FastAPI
• PostgreSQL
• Redis
• Kubernetes
• WebSockets
• Streaming systems
• ASR, TTS, Vision, LLM model APIs
• FFmpeg
• pyav
• VAD (voice activity detection)