AI Agent Developer – PriceLabs Revenue Manager for Vacation Rental Portfolio

We operate a professional short-term and mid-term rental management business and are building a portfolio-level AI Revenue Management Agent integrated with the PriceLabs Customer API. This is not a chatbot project. The system will monitor multiple PriceLabs listings, analyze property performance against appropriate market benchmarks, audit pricing settings, identify revenue opportunities, explain its recommendations, and eventually make controlled PriceLabs changes through an approval workflow. The goal is not simply to maximize occupancy. The agent should help maximize sustainable revenue and RevPAR while outperforming the appropriate market and comparable properties. Phase One: Read-Only Shadow Mode The first contracted phase will operate in READ-ONLY SHADOW MODE. The agent must analyze our real portfolio, generate and log recommendations, track subsequent results, and determine what it would have changed without actually modifying live PriceLabs pricing or settings. The system must support: * Multiple properties within one portfolio * Different strategies, markets and constraints for individual properties * Property-specific economic floors and owner requirements * Portfolio-wide daily monitoring * Prioritization of the most important revenue opportunities * Market and comparable-property performance analysis * PriceLabs settings audits * Explainable recommendations * Complete decision and audit logging * Onboarding newly added PriceLabs listings without requiring source-code changes The system should be designed so that later phases can add owner-approved PriceLabs changes and, eventually, limited autonomous actions after the agent has demonstrated sufficient reliability. Detailed Specification Provided We have already created a detailed developer specification covering: * System architecture * PriceLabs integration * Portfolio-level operation * Revenue-management decision logic * Property-specific profiles * Green / Yellow / Red permissions * Deterministic safety guardrails * Shadow mode * Human approval workflows * Testing requirements * Acceptance scenarios * Security * Logging and explainability * Development milestones * Future expansion Please review the attached PDF before submitting your proposal. Required Technical Experience Candidates should have strong experience with: * REST APIs and API integrations * Python and/or TypeScript/JavaScript * AI agents and LLM tool/function calling * Structured LLM outputs * Deterministic business rules and guardrails * Databases and structured data * Scheduled/automated workflows * Audit logging * Authentication and secrets management * Human-in-the-loop approval systems * Error handling * Testing * Production deployment Experience with PriceLabs, vacation rentals, property-management systems (PMS), hospitality technology, dynamic pricing, or revenue-management systems is highly desirable. Please Include the Following in Your Proposal 1. Examples of AI agents, API-driven automation systems, or similar applications you have personally built. 2. Your proposed technical stack for this project. 3. How you would separate deterministic business rules and safety controls from LLM reasoning. 4. How you would prevent the AI agent from making unauthorized PriceLabs changes. 5. How you would implement read-only shadow mode. 6. How you would log recommendations, decisions and eventual actions so they can be audited later. 7. Your testing strategy before any PriceLabs write access is enabled. 8. How you would architect the system so that additional properties can be added without rewriting the software. 9. Your estimated timeline. 10. Your fixed-price estimate for Milestones 1–4 described in the attached specification. Please quote later write-access/production milestones separately. Important Screening Instruction Begin your proposal with the words “Shadow Mode.” Then, after reviewing the attached specification, identify the biggest technical challenge you anticipate in building this system and briefly explain how you would address it. This instruction is part of our screening process. Generic proposals that do not address the attached specification will not be considered. We are looking for a developer who can build a reliable, secure and maintainable production system, not simply demonstrate an AI prototype. This project may lead to additional AI-agent development for other areas of our property-management business if the initial project is successful.