AI Research Engineer

Location

In person, Central London, 4-5 days in office

Type: Full-Time

NB: We are able to sponsor visas

Note on Salary: For exceptional candidates with strong product impact and relevant experience, we are open to offering compensation above the specified range.

Arva AI is revolutionising financial crime intelligence with our cutting-edge AI Agents. By automating manual human review tasks, we enhance operational efficiency and help financial institutions handle AML reviews, while cutting operational costs by 80%.

As an AI Research Engineer, you’ll play a pivotal role in building our in-house models—training, fine-tuning, and rigorously evaluating LLMs and agentic systems that power our AI compliance platform, from document fraud detection to web-scale due diligence. You’ll also develop the automation and infrastructure that continuously tests and improves performance (evals, data/label pipelines, training loops, and iterative fine-tuning/distillation).

About the Role

As an AI Researcher, you will

• Build and iterate on in-house LLM and agentic capabilities, integrating them into product features with rigorous evaluation, benchmarking, and backtesting across real compliance workflows.
• Design continuous-improvement loops for our models, leveraging human-in-the-loop feedback to drive dataset curation, fine-tuning/distillation, and optimisation of model behaviour.
• Stay on the frontier and translate research into impact, exploring and applying new AI methods to improve model quality, explainability, latency, and reliability in production.

What You’ll Do

• Evals: Build benchmarks, evaluate backtests and analyse results to quantify our performance and prevent regressions.
• Model Fine-Tuning: Adapt and fine-tune large language models and vision models for specific use cases.
• Custom Model Training: Train and deploy bespoke AI models tailored to solve unique compliance challenges.
• AI features: Use the latest in LLMs, agents and computer vision to build systems that are faster, more thorough and more accurate than humans.
• MLOps: Ensure that we maintain good practices around collecting and managing datasets, backtests and benchmarks to to maximise development velocity
• Prompt Engineering: Develop and test prompts to optimise model performance and deliver high-quality outputs.
• Collaboration: Work closely with the engineering and product teams to translate customer needs into actionable AI solutions.
• Iterate: Continuously iterate on our AI systems to improve performance, reduce errors, and deliver value to customers.

Our Culture

Deliver Value Fast

Speed starts with clarity. We first understand what value actually means, for the customer, the business, or the system, and then take the shortest credible path to delivering it. We move quickly without cutting corners on quality, security, or trust.

Outcome Obsessed

We obsess over details and take full ownership of wider outcomes, not just tasks. We think like the user, execute with rigor, and validate that every detail works in real conditions. If something falls short, we fix it properly and prevent recurrence, always raising the bar.

Relentless Urgency

We move with urgency because time matters. We prioritise what truly moves the outcome, make decisions with imperfect information, and act decisively. Urgency means focus, momentum, and driving meaningful progress without unnecessary delay.

What We’re Looking For

Experience

• 3+ years in an AI research or engineering role, with experience building and testing agentic systems.
• Technical Expertise: Hands-on experience with prompt engineering, fine-tuning pre-trained models, and training custom models, including vision models. Experience using TypeScript a plus.
• Product Mindset: Strong understanding of how AI can solve real-world problems, with a focus on customer needs.
• Interests: Ability to stay updated with the latest advancements in AI and apply them effectively.
• Collaboration: Excellent communication skills to work across teams and explain com