Head of Research (£130k-£180k + Equity) at Arva AI

Job Title

Head of Research

Salary

£130k-£180k + Equity

Company Description

Arva AI is a YC-backed (S24) startup, funded by Gradient Ventures (Google's AI fund), building specialized AI agents that automate complex financial crime compliance for banks and global fintechs.

Job Description

As Head of Research, you will own Arva AI's AI strategy, roadmap, and model quality outcomes. You will lead a team building agentic systems that handle 92% of financial crime reviews automatically. Your work directly enables regulated institutions to trust AI for high-stakes AML and KYC workflows through rigorous evaluation and explainability.

Location

London, UK

Why this role is remarkable

• Lead the research function from the ground up, setting the technical agenda and building a world-class team for a fast-scaling YC startup.


• Join a well-funded team (seed led by Gradient Ventures) already shipping product to regulated financial institutions with real-world impact on global compliance.


• Solve unique challenges at the frontier of applied AI, where model accuracy, speed, and explainability are critical for regulatory trust and automation.

What You Will Do

• Define the research roadmap, making strategic build-vs-adapt decisions across fine-tuning, distillation, and bespoke model training for LLMs and vision systems.


• Own company-level model quality, establishing the benchmarks, evaluation infrastructure, and regression suites required for high-accuracy financial crime automation.


• Partner cross-functionally with product and engineering teams to bridge the gap between research and production, ensuring rapid, safe deployment of new models.

The ideal candidate

• Has 5+ years of experience in applied AI research or ML engineering, with a proven track record of moving research agendas into production.


• Possesses hands-on expertise in training and fine-tuning LLMs and vision models, specifically within agentic systems or high-reliability environments.


• Thrives in early-stage startup environments, demonstrating comfort with ambiguity and a strong desire to own mission-critical technical outcomes from day one.