AI Engineer

Ryz Labs is looking for an AI Engineer to work with one of our clients. The role This is an AI engineering role with a forward-deployed edge. You'll split your time between: 1. Building our core AI systems: the agents, pipelines and evaluation infrastructure behind everything Smart Bricks does 2. Deploying them with clients: working embedded with real estate private equity funds, institutional investors and family offices to solve their hardest problems with AI You write production code, ship it and own it. On client work you partner with our GTM team: they open the door and frame the problem, you scope, build and deploy the solution. What you learn in the field goes straight back into the platform, so every deployment makes the product better.   What you'll do Core AI engineering • Build agentic systems: multi-step agents, tool use and workflows that handle real investment work: sourcing, underwriting, due diligence, IC memos and portfolio monitoring • Own the AI stack: RAG and document pipelines, model routing across frontier and open models, prompt and context engineering, structured outputs into Excel and IC formats • Make it trustworthy: build evals and observability for accuracy, cost, latency and reliability. In finance, every number has to trace back to source • Build reusable infrastructure: connectors, data pipelines and components that make each new deployment faster than the last Forward-deployed • Embed with clients: learn their data, workflows and decisions well enough to build for them • Partner with GTM: join discovery and scoping calls, turn messy client problems into a technical plan, and help close and expand accounts • Prototype, validate, ship: get something real in front of the client within days, iterate on their feedback, and take it to production • Own the technical relationship: be the person the client's team trusts, from first deployment through to rollout • Feed the platform: turn one-off client builds into product features What you bring • 3+ years shipping production software, ideally in Python and/or TypeScript • Deep hands-on experience building with LLMs in production: agents, RAG, tool use, evals and context engineering (LangGraph, LlamaIndex, MCP or similar) • Strong grasp of the AI landscape: which models to use for what, and how to trade off accuracy, cost and speed • Solid ML fundamentals: evaluation, problem decomposition, and knowing when not to use an LLM • A builder's track record: founding engineer, early startup hire, ex-founder or ex-CTO, or a top engineering team where you owned outcomes end to end • Comfortable in front of clients. You can explain technical trade-offs to a fund partner, then go and build it • Fast and pragmatic. You'd rather ship something real this week than perfect a diagram • Cloud fundamentals (AWS, GCP or Azure), data pipelines, and awareness of enterprise security: data isolation, access controls, no training on client data • Comfortable working across UK and GCC hours