Role : Senior Software Engineer AI Platforms.

Location : Pune.

Experience : 4+ year.

About the Role :

We're building an agent management platform that automates our software development lifecycleand eventually all organizational processes across engineering, product, and marketing.

You'll design and build the infrastructure that orchestrates AI agents to transform how work gets done.

This isn't about using AI as a tool. It's about building the system that manages AI agents at scale.

What You'll Do :

- Architect and build our agent orchestration platform from the ground up.

- Design automation workflows that coordinate multiple AI agents across business functions.

- Implement evaluation frameworks and test harnesses for LLM-based systems.

- Build reliable, deterministic pipelines on top of probabilistic AI outputs.

- Create the primitives and abstractions that other teams use to deploy agents.

Required Qualifications :

- Core Engineering (4+ years).

- Strong Python fundamentals; TypeScript experience valued.

- System design for distributed, async workloads, API design and integration patterns.

- Database design for AI/ML workloads (vector stores, event sourcing) LLM Engineering.

- Deep understanding of LLM primitives : prompts, context windows, tool use, structured outputs, embeddings.

- Experience with agentic frameworks (LangChain, LangGraph, CrewAI, AutoGen, or similar).

- Familiarity with AI-assisted development tools (Claude Code, Cursor, Copilot, Aider).

- Understanding of RAG architectures and retrieval patterns.

Reliability & Testing for AI Systems :

- Test-driven development practices adapted for non-deterministic systems.

- Experience building evaluation frameworks for LLM outputs (accuracy, latency, cost).

- Techniques for making AI workflows reproducible and debuggable.

- Understanding of guardrails, validation, and human-in-the-loop patterns.

Preferred Experience :

- Built production systems that orchestrate multiple AI agents.

- Implemented CI/CD pipelines that include LLM evaluation steps.

- Experience with MCP (Model Context Protocol) or similar tool-use standards.

- Contributed to or used agent frameworks for workflow automation.

- Familiarity with cost optimization for LLM-heavy workloads.