This is a job posting for a Forward Deployed Engineer focusing on backend Python for a funded and profitable stealth startup in property management. The role involves building AI agent backend tools, handling client-specific extensions, and direct client interaction to solve real-world problems with AI agents. It's a senior role requiring experience with production backend systems and LLM agents.
A starter prompt for Claude Code, what you'll need, and how to reach them.
You are a senior backend Python engineer tasked with building an AI agent's tool-calling backend for property management. The primary goal is to create a small, demonstrable system where an AI agent can handle maintenance requests. Focus on a single use case: receiving a text description of a maintenance issue, classifying it (e.g., 'plumbing', 'electrical', 'HVAC'), and then calling a mock tool to 'dispatch' a technician. Your system should: 1. Set up a Python backend using FastAPI. 2. Define a simple REST API endpoint to receive maintenance requests. 3. Integrate with a local LLM (e.g., via `ollama` or `liteLLM` for a local setup, or a mock LLM for quick iteration) to classify the request. 4. Implement a 'tool' function (e.g., `dispatch_technician(issue_type, description)`) that the LLM can call. For the MVP, this tool simply logs the dispatch action. 5. Include basic error handling and logging. 6. The output should be a working Python project, with `requirements.txt` and clear instructions on how to run it. Emphasize clean, production-ready code. The final deliverable is a functional backend service that processes a maintenance request through classification and tool-calling by an AI agent.
Reach out to this team to offer collaboration on agent evaluation tooling with 'agent-eval-lab' or provide consulting on building robust agent platforms using principles from 'mcp-kit', emphasizing reliable AI operations.
Forward Deployed Engineer - backend - python | REMOTE (UTC -3 to +3) | Full time Small stealth team in the property management space, building AI agents that run real operations, where agents and humans work as one team rather than as user and tool. Funded, profitable, fully remote, competitive comp. We want a senior backend engineer to own a client facing use case end to end: build the backend tools and workflows our agents call, handle the jurisdiction and client specific extensions, and talk to the client directly. Pull requirements, explain the tradeoffs, own the relationship. Take messy real world problems, turn them into something agents can actually use, then keep it alive in production. You're probably a fit if: - you've shipped serious backend systems in production - you can hold your own on a customer call - you've run LLM agents in production, or you'll ramp fast: tool calling, evals, knowing when a plain deterministic tool beats a model call Bonus, not required: containers and CI/CD, RBAC and auditing, LangChain and LangGraph, prior customer facing or operator background. We care more about what you've built and can build than a checklist, so skip the skills inventory a
Email mercedes@kinxshn.com with "HN - Forward Deployed Engineer" in the subject line.
“I'm a solo operator with deep experience in shipping robust backend systems and building AI agents in production. I've successfully developed and maintained systems that bridge real-world problems with agent capabilities, including strong evaluation frameworks. Here's a link to a demo project that illustrates my ability to build an end-to-end AI agent tool-calling backend for a property management-like use case, focusing on classification and deterministic tool execution.”
Open the original ↗