I need to be direct: I'm a CEO with deep regulatory networks across UK financial services, insurance, and AI policy — King's Counsel, ex-FCA regulators, Lloyd's underwriters. What I don't have is a technical co-founder who can build the infrastructure that makes those networks worth billions.Here's the unsolved problem:$650B+ is being committed to sovereign AI globally. The EU AI Act is enforcing. UK SM&CR makes executives personally and criminally liable for AI they deploy. Lloyd's is excluding GenAI from policies entirely. Every regulated institution deploying AI needs independent verification — and no infrastructure-grade solution exists.We're building the verification layer. Not compliance SaaS. Not an LLM wrapper. Think what Moody's built for credit ratings — infrastructure that sits between AI systems and the regulated world. Harvey AI (Sequoia, GV, $5B valuation) is doing this for legal. Vanta (Sequoia, $4.15B) did it for compliance. We're doing it for sovereign AI verification.The hard research problems that make this interesting:→ How do you build verification models for AI systems that are themselves evolving? Static audits don't work for dynamic systems. This requires novel approaches to continuous evaluation.→ How do you design RLHF annotation schemas that capture regulatory nuance across jurisdictions — UK SM&CR, EU AI Act, Lloyd's exclusions — without collapsing domain expertise into generic labels?→ How do you create a data flywheel where every expert attestation (from lawyers, ex-regulators, domain specialists) makes the next automated review more accurate — and prove that accuracy to regulators?→ How do you build sovereign-first infrastructure with structural immunity to foreign jurisdictional overreach (CLOUD Act) while maintaining production-grade reliability?→ How do you architect real-time telemetry for multi-agent AI systems with sub-second intervention capabilities?You'd be building this from zero as sole engineer in Year 1. Designing the architecture. Shipping the MVP. Then scaling the team as revenue grows (bootstrap model — each phase funds the next). Title evolves to CTO.And you'd be in every investor meeting. We're targeting Sequoia, GV, and Tier 1 VCs. You need to explain why the sovereign data moat, the RLHF flywheel, and the infrastructure lock-in create a defensible, category-defining company.This is a genuine co-founding role:• Significant equity (4-year vest, 1-year cliff, milestone acceleration)• Full-time, London-based• Below-market cash Year 1, scaling with revenue• Targeting profitability by Month 9, major VC round Year 2• Every £1B in sovereign AI investment creates £10–25M in governance demand — the TAM is structurally growingYou might be:→ A researcher or engineer from DeepMind, Google Brain, Meta FAIR, or a frontier AI lab who wants to apply ML to a genuinely consequential domain where the data is proprietary and the moat is real — not another chatbot.→ A senior ML engineer who's built production RLHF, fine-tuning, or annotation pipelines and wants co-founder equity in a company where their work creates the core defensible asset.→ An exceptional Cambridge, Imperial, Oxford, or UCL CS graduate who's been building since undergrad — open source, shipped products, published research — and wants to go from zero to CTO of an infrastructure company.Must-haves: You write production code today. Python + cloud infrastructure minimum. You can explain technical trade-offs to a non-technical investor without losing credibility. You're in London or relocating. Full-time from day one.DM me with:(1) What you've built or researched — be specific(2) A link (GitHub, paper, product, anything)(3) Why this particular problem interests you(4) Your availabilityNo CVs. Builders and researchers only.#CoFounder #FoundingEngineer #DeepTech #AIGovernance #RLHF #London #Hiring #InfrastructureEngineering #MLEngineering