Software Engineer (AI Agents)

ProdE forms org-wide codebase intelligence - helping teams plan, build, and ship the software the world runs on. We outscored DeepWiki by 15%, Google Code Wiki by 38%, and Claude Code by 40% on AI codebase documentation benchmarks.

Now we're building AI agents for one of the hardest problems in enterprise software: planning. The systems you build will reason about high-stakes decisions for billion-dollar enterprises, where being wrong is expensive and being right is transformative.

This is not a wrapper-around-an-API job. You'll write production-grade code that holds up at scale and design agents that work when reality gets messy.

What you'll do

• Design, build, and ship production AI agents for enterprise planning.
• Make deliberate architectural choices: planning loops, tool use, memory, multi-agent, human-in-the-loop - and know when not to use each.
• Write production-grade Python that scales and stays maintainable.
• Build supporting infrastructure: data pipelines, retrieval, evaluation harnesses.
• Iterate on reliability, accuracy, and cost.

What we're looking for

• 2+ years of professional software development.
• Strong Python with a track record of production code at scale, not just prototypes.
• Hands-on experience building AI agents, not just using them.
• Real understanding of where LLM-based agents excel and where they fail.
• Familiarity with ReAct, planning/reflection loops, tool use, multi-agent orchestration, RAG-augmented agents, and the tradeoffs.
• MongoDB: schema design, queries, production use.

Nice to have

• Agent frameworks (LangGraph, LlamaIndex, CrewAI, AutoGen) - and comfort without them.
• Background in planning, scheduling, optimization, or operations research.
• Evaluation and observability tooling for non-deterministic systems.
• Enterprise software exposure.

Who you'll work with

A small, senior team - the co-founders, every day. Abhishek (CEO) led AI Agents at Leena AI (YC S18) from $100K to $10M ARR. Nilesh (COO) was a Director at a Series B company and learned ML at CMU. Our mentor, advisor, and investor Mitz Banarjee backs Anthropic, SpaceX, xAI, Perplexity, Groq, Cerebras, and Figure - and helped take Workiva (NYSE: WK) from founding to IPO.

How we work

• Core team, real ownership, early-team ESOPs on the table.
• Remote-first. Gurugram preferred for occasional in-person.

Important
In your applications, highlight your best work - not something vibe-coded from a single prompt, but real ingenuity and product thinking on a hard problem.