We build AI agents that run revenue for short-term rental and hotel operators on Booking.com and Airbnb: live prices, promotions and visibility on real portfolios, plus an MCP connector so customers work their accounts from Claude and ChatGPT.
Our engineers no longer type most of the code. Claude Code and Codex do. The hard part is knowing what to build, how it should fit together, and whether what the agent wrote is actually right. You will talk to customers, own the architecture, write the briefs, and catch what the agent missed.
We do not care which language you know.
We care that you can do architecture (single source of truth, clear boundaries, the right layer for a fix) and that you already ship this way.
Big bonus if you have worked on hotels, short-term rentals or dynamic pricing.
Small team, paying customers, hard problems (private OTA APIs, PMS integrations, pricing decisions that move real money). Three interview stages, no take-home. Email me (CTO) directly: [email protected]
We're an AI agent that automates pricing and promotions for short-term rental operators on Booking.com and Airbnb. Bootstrapped to real revenue, recently funded, small team.
Pitch: we sit on Booking.com private API data, Airbnb channel data, and scraped competitive pricing across thousands of properties.
Our agent is rules-based where it should be data-driven. You'd be the founding ML/data hire fixing exactly that.
You're a fit if you:
* Ship clean production Python in a Django codebase, not just notebooks
* Think in experiments and know correlation from causation (time series a plus)
* Use frontier AI coding tools/agents to ship fast
* Have startup experience and thrive in ambiguity
Stack: Python, pandas, scikit-learn, PostgreSQL, ClickHouse, Django, time series, causal inference, AI agents.
Email me at [email protected].
https://mydatavalue.github.io/polter/
I've built Polter: https://github.com/myDataValue/polterAn open source React library that lets AI agents execute tasks on your product by using your actual UI instead of doing API calls.Why? Because this solves the issue for many apps where users don't understand how to use your product, most agents are just API calls right now.This way the user can understand your UI implicitly reducing...