AI/ML Engineer

Job description

About the role Globixs is hiring an AI/ML Engineer to own model development and agent orchestration inside our Real Estate product line. You'll work directly with the AI Architect on the next phase of the platform: expanding our XGBoost-based Automated Valuation Model (currently R² = 0.95), shipping the Property Intelligence Module to paying brokerages, and integrating live MLS and Zillow data. This is a builder role. You'll ship to production, not prototype in notebooks. What you'll do - Own and improve the AVM pipeline (XGBoost today; open to LightGBM, CatBoost, or neural approaches where they win) - Build and extend our multi-agent property intelligence system (PropertySearchAgent, MarketAnalysisAgent, PropertyValuationAgent) using the Claude and OpenAI APIs - Integrate third-party data sources — RapidAPI Zillow, MLS feeds, county assessor data, OpenCage, US Census — into clean, versioned feature pipelines - Design prompts, tool definitions, and orchestration logic for agents running on FastAPI + Supabase + Railway - Ship features end-to-end: data ingestion → model → API → the realtor-facing dashboard - Monitor model drift, evaluate agent outputs, and build eval harnesses for LLM-driven components - Collaborate with frontend engineers on Next.js integrations and with the founder on customer-driven priorities Tech stack Python · FastAPI · XGBoost · Supabase (Postgres) · Claude API (Haiku + Sonnet) · OpenAI · Make.com · n8n · Railway · GCP · Next.js / Streamlit · GitHub Actions How to apply Send to connect@theteammc.com: 1. Resume or LinkedIn 2. A link to something you've shipped (GitHub, live demo, or write-up) 3. One paragraph: what would you change about a typical Zillow Zestimate, and why? We review on a rolling basis and reply to every applicant who answers #3.

Minimum qualifications

• 2–5 years building ML systems in production (not just notebooks)
• Strong Python; comfortable with FastAPI, Pandas, scikit-learn, and a gradient-boosting framework (XGBoost or LightGBM)
• Hands-on experience with LLM APIs (Anthropic Claude and/or OpenAI) — prompt design, tool/function calling, structured outputs
• Experience with at least one multi-agent framework or custom orchestration pattern (LangGraph, CrewAI, OpenAI Agents SDK, or hand-rolled)
• Solid SQL and Postgres; familiar with Supabase or similar BaaS
• Git-based workflows, CI/CD (GitHub Actions), and cloud deploys (Railway, GCP, Vercel, or AWS)
• Ability to read a product spec, ask the right questions, and ship
• Nice to have: real estate / proptech domain experience (MLS, AVM, RETS, comp analysis)
• Nice to have: geospatial work (PostGIS, H3, shapely, OSM)
• Nice to have: Streamlit or React for demo-ing models to non-technical users
• Nice to have: LLM evals / observability (Langfuse, Braintrust, or home-grown)

Preferred qualifications

• Client-facing delivery and cross-functional collaboration experience.
• Strong communication and stakeholder management skills.
• Experience owning end-to-end delivery outcomes.