Sarvam is building India's sovereign AI platform, focusing on research, models, infrastructure, and applications to make AI work for India. We partner with leading enterprises and public institutions, including Tata Capital, SBI Life, and CRED. Our conversational AI platform, Samvaad, empowers enterprises, startups, and governments to build, configure, deploy, and operate voice agents at scale.

This platform currently handles millions of personalized voice calls daily, demonstrating deep user understanding. We are seeking a Staff Engineer to take end-to-end ownership of this critical system, encompassing its data model, APIs, orchestration, reliability, and performance standards. This is an individual contributor role focused entirely on technical systems and excellence, without people management responsibilities.

Take full ownership of the backend systems for agent, campaign, and user management, from data modeling and API design through to deployment, utilizing Python and FastAPI.

Implement workflow orchestration using Temporal for complex, long-running, stateful operations.

Ensure high performance and reliability for a multi-tenant SaaS platform, adhering to sub-second latency Service Level Objectives (SLOs).

Develop robust observability solutions, including logging, metrics, and tracing, to enable confident and rapid debugging of production issues.

Build and maintain integration testing infrastructure to facilitate fast iteration without compromising customer-facing systems.

Collaborate closely with the frontend team (Next.js) to deliver a seamless user experience that complements the underlying platform.

Champion a design-first and documentation-first engineering culture, emphasizing RFCs before code, documented decisions, and the elimination of tribal knowledge.

Possess 8-12 years of experience building production backend systems, with deep expertise in Python and FastAPI (or Golang).

Demonstrated ability to orchestrate high-volume, personalized campaigns and long-running conversations with strong reliability.

Proven track record of full-stack ownership, including designing systems end-to-end, making critical trade-offs, and operating them in production.

Hands-on experience with workflow orchestration at scale, ideally using Temporal.

Experience building and operating high-scale distributed systems, including multi-tenant, low-latency, and resilient architectures under real load.

Strong foundational knowledge of PostgreSQL and Redis, including schema design, query performance optimization, and caching strategies.

Excellent judgment regarding build vs. buy decisions, optimization vs. shipping priorities, and abstraction vs. inlining approaches.

Experience building products on top of LLMs or AI/ML systems is a plus.

Familiarity with telephony, VoIP, or SIP infrastructure is advantageous.

Previous experience in an early- or growth-stage startup, demonstrating comfort with ambiguity, speed, and ownership, is desirable.