Oun Homes is turning the chaotic contract-to-close phase of a real estate deal into a self-driving workflow. We’re building an LLM-powered email agent that reads every message, understands and prepares every document, and takes the next best action—without asking the agent to open “yet another platform.&quot;Behind the scenes, we pair LLM reasoning with a document intelligence pipeline that can classify, extract, and prepare documents for legally binding signatures. No new log-ins, no new habits—just deals that move themselves forward.We’ve raised pre-seed capital from top VCs &amp; angels in NYC and the Silicon Valley.

Oun Homes

About Oun Homes (ounhomes.ai):

- Oun Homes is turning the chaotic contract-to-close phase of a real estate deal
into a self-driving workflow. We’re building an LLM-powered email agent that
reads every message, understands and prepares every document, and takes
the next best action, without asking the agent to open “yet another platform”
- Behind the scenes, we pair LLM reasoning with a document intelligence
pipeline that can classify, extract, and prepare documents for legally binding
signatures. No new log-ins, no new habits—just deals that move themselves
forward
- We’ve raised pre-seed capital from top VCs & angels in NYC and the Silicon Valley


The Role:
We’re looking for an AI Engineer to help develop its brain, memory, and toolbelt.
You’ll build intelligent systems that fuse LLMs, document-vision tech, and agentic
workflows to automate the complexities of real-estate transactions.

As an AI Engineer, you will:
- Ship end-to-end AI features: Own the entire stack—from data pipelines and model orchestration to the React front-end—turning ideas into production-ready code in days, not weeks
- Design the agent’s cognition: Architect multi-step, tool-calling workflows—
retrieval, reflection, planning, execution, and self-evaluation loops
- Build a robust memory layer: Stand up vector-search and structured stores so
the agent remembers every email, doc, and decision
- Integrate real-world tools: Wire the agent into email, DocuSign, and our
document-intelligence pipeline to close the loop
- Raise the agent’s IQ: Own prompt engineering, function schemas, guardrails,
and evaluation
- Iterate at startup velocity: Rapidly prototype, measure, and refine
experiments; push improvements on a tight feedback loop so systems keep
getting smarter with every commit
- Influence everything: Setting the hiring bar, engineering culture, product
roadmap, and technical vision

In This Role, You Will Thrive If You:
- Have 3+ years shipping production software (bonus points if 0 → 1 at a
startup)
- You’re experienced across the stack, proficient in Python for backend and
TypeScript & React for frontend
- You’ve built or maintained systems using frameworks like LangChain,
LlamaIndex, or similar, including RAG, function calling, and tool execution
- You’ve deployed pgvector, Pinecone, Elastic, or similar, understand
embedding pipelines, and can tune similarity search for real workloads
- You’re a self-starter with high learning velocity—if a library, framework, or
domain is unfamiliar, you can ramp up quickly
- You’re eager to work with cutting-edge tools to build frontier AI systems
- Experience in real estate or document processing is a plus

*Outstanding performance may convert into a core team
role*