Agentic Platform Architect

Reliance Intelligence is the deep-tech AI venture of Reliance Industries, on a mission to make Artificial Intelligence accessible, trusted, and affordable for 1Bn+ Indians. We're not bolting AI onto legacy infrastructure — we're building India's sovereign AI backbone from first principles: gigawatt-scale, powered entirely by clean energy, and engineered to be one of the largest AI platforms anywhere in the world.

On this foundation, our AI services will work in 22 Indian languages and span consumer, enterprise, devices, and social good built around one principle: AI must be easy to use, trusted to rely on, and affordable for all. We're building a world-class team of researchers, engineers, PMs, and technology leaders, and partnering with global technology players, universities, and research institutions to shape the future of intelligence, for the world, from India.

The Role

We're looking for an Agentic Platform Architect (Shop OS & Runtime) to design how agents run, coordinate, and execute inside Shop OS. You'll work closely with the Product, Data, DevOps, and QA pods within your vertical to run a stable agent execution runtime, with deterministic + LLM hybrid workflows, safely across millions of shops.

This is an AI-native, hands-on, 0-to-1 build organization. Everyone here is hands-on — there are no manager-only seats — and roles differ by problem focus, not by SDLC phase ownership. There's no handoff between Product, Engineering, QA, Data, and Ops: you're expected to design, build, test, deploy, and observe within your own domain, comfortable with ambiguity every step of the way.

What You'll Do

• Architect a stable agent execution runtime with deterministic + LLM hybrid workflows.
• Design cost-optimized inference orchestration at India scale.
• Own systems architecture and observability for agents running across millions of shops.
• Write code daily and set security best practices across the platform.
• Partner with vertical leads to scale systems with maximum stability.

What We're Looking For

• 10+ years in distributed systems or platform engineering, having built production AI platforms (not POCs.)
• Hands-on experience building with LLMs / AI agents / ML systems in production (not just prototyping.)
• Strong Java/Node/Python, event-driven architecture (Kafka), vector DB/RAG/tool-calling, and workflow engines (Temporal.)
• Comfort with ambiguity and a fast-moving, 0-to-1 environment — we ship weekly, not quarterly.
• Bonus: open-source contributions, and a strong focus on real user monitoring and infrastructure observability.