Production AI Agents in Finance - AAIF London x Gradient Labs

How are teams actually building, evaluating and operating AI agents in high-stakes financial environments?

Join AAIF Community London, in partnership with Gradient Labs, for an intimate technical evening on production AI agents in financial services and fintech, including the engineering systems required to operate them reliably.

Four focused talks will examine real systems and engineering trade-offs across evaluation, reliability, agentic product design, infrastructure, developer tooling and AI-native software engineering. The talks will be followed by a speaker panel and audience Q&A moderated by Gleb Lukicov, MLOps Lead at CIBC Capital Markets.

The first hour is for food, drinks and networking.

Capacity is limited to approximately 50–60 people. Please register only if you intend to attend.

Speakers & Talks

Ibrahim Faruqi - AI Engineer, Gradient Labs

Building Agents You Can Bank On in Production

Gradient Labs’ agents are trusted and used by multiple banks and fintech organisations in high stakes production workflows including customer support. We’ll explore how we evaluate and benchmark our agents in these complex, regulated environments, to build and maintain that trust.

Bio: Ibrahim is an AI engineer at Gradient Labs, where he’s spent the last 2 years building AI agents across text and voice for financial services. Before Gradient Labs, he spent four years at Monzo building production ML systems for financial crime, fraud detection and personalised recommendations.

Marlin Ranasinghe - Senior Machine Learning Engineer, Starling

Building the UK’s First Agentic AI Assistant

Marlin will dive into the technical challenges behind building the Starling Assistant, moving beyond chat to an AI that takes action on behalf of customers. He’ll share key lessons learned across tool design, tool confirmations, mobile-native agentic UI, and tool versioning, before wrapping up with what’s next for agents across industry.

Bio: Marlin is a Senior Machine Learning at Starling, focused on foundational AI, with interests in agentic AI, inference and post-training.

Dominic Black - Backend & Platform Engineer, Morse

Self-Healing Is a Routing Problem

Its own tickets, worked by its own agents: how Voluntold builds and fixes itself.

Voluntold is the routing layer between your boards and the coding agents that you already run, on your machines using your accounts. The same loop that takes tickets off your boards to implement them also takes our tickets, filed by firing alerts, and bugs our customers' coding agents find in Voluntold. This talk is about that loop in production, and where it stops when the fix is wrong.

Bio: Dom is a Backend & Platform Engineer at Morse, previously a Staff Engineer at Monzo and Founding Engineer at Encore.dev. He has spent 15 years building ledgers and payment rails in regulated fintech, and sat on the Bank of England’s CBDC Technology Forum.

Ibrahim Malik - AI/ML Engineer · Technical Reviewer, O’Reilly

AI-Native Software Engineering in Finance

Building Financial Software with AI Agents

AI-native software engineering is not simply about generating code faster. Drawing on the forthcoming O’Reilly book AI-Native Software Engineering: Building Software with AI Agents and Spec-Driven Development, this talk explores how specifications, context engineering, harnesses, orchestration and verification reshape the way teams build financial software with AI agents, with a particular focus on review, feedback and quality gates. Through a practical example, Ibrahim will examine why plausible output is not enough in high-stakes systems and what remains essential for producing dependable, accountable software as models improve.

Bio: Ibrahim Malik is an AI/ML engineer and AAIF London Lead focused on agentic systems, evaluation and production AI. He is a technical reviewer for O’Reilly’s forthcoming AI-Native Software Engineering, with recent work spanning trace-first release gates and guarded multi-agent workflows.

Panel Moderator

Gleb Lukicov - MLOps Lead, CIBC Capital Markets

Gleb leads MLOps and agentic AI strategy and implementation at a global investment bank. He will bring the four speakers together for a discussion about what it takes to move agents from promising prototypes to dependable systems in high-stakes financial environments.

Agenda

17:30 - 18:30 | Doors Open, Food, Drinks & Networking

Check in, grab some food and drinks, and meet engineers, researchers, founders and practitioners working across AI agents, financial services and production infrastructure.

18:30 - 18:40 | Welcome & Event Introduction

Opening remarks from Alex Jones and Julie Huang, introducing AAIF Community London and the evening’s technical theme.

May will introduce each speaker immediately before their talk.

18:40 - 18:55 | Ibrahim Faruqi - Gradient Labs

Building Agents You Can Bank On in Production

18:55 - 19:10 | Marlin Ranasinghe - Starling

Building the UK’s First Agentic AI Assistant

19:10 - 19:20 | Break

Grab a drink and continue the conversation.

19:20 - 19:35 | Dominic Black - Morse

Self-Healing Is a Routing Problem

19:35 - 19:50 | Ibrahim Malik - Technical Reviewer, O’Reilly

AI-Native Software Engineering in Finance

19:50 - 20:25 | Speaker Panel & Audience Q&A

Moderated by Gleb Lukicov, MLOps Lead at CIBC Capital Markets.

The panel will bring the speakers together around questions including:

• What evidence is sufficient to trust an agent in a high-stakes workflow?
• Where do current evaluation methods create false confidence?
• How should teams decide what an agent may do autonomously and when it must escalate?
• What changes when an agent moves from answering questions to using tools and taking consequential actions?
• Which infrastructure and engineering practices will remain important as models improve?

We’ll then open the discussion to questions from the audience.

20:25 - 21:00 | Networking

Continue the conversation with the speakers and attendees.

21:00 onwards | Optional Social

For anyone who wants to continue the evening, we’ll head somewhere nearby.

Location shared on the night.

Who Should Come

This event is for AI/ML engineers, research engineers, quantitative researchers, software engineers, agent developers, applied AI teams and infrastructure engineers working on or interested in:

• production agentic systems
• financial services and fintech
• agent evaluation and reliability
• AI-native software engineering
• agent infrastructure and developer tooling
• post-training and feedback loops
• high-stakes applied AI

If you’re building agents - or the infrastructure and software around them - and thinking seriously about how they move from prototypes to dependable production systems, the discussions should be relevant.

About Gradient Labs

Gradient Labs builds AI agents for financial services.

Founded by early Monzo engineers, the team is working on production agent systems for financial institutions, where reliability, control and operational performance matter.

Gradient Labs is supporting the event with the venue, food and drinks.

​About AAIF

The Agentic AI Foundation (AAIF) is a Linux Foundation initiative supporting an open and interoperable agentic AI ecosystem.

AAIF brings together developers, researchers, open-source maintainers and organisations working across agent protocols, infrastructure, tooling and interoperability.

AAIF Community London runs technical events across agentic AI, machine learning, developer tooling and AI infrastructure.

By attending, you agree to the Linux Foundation Code of Conduct and Privacy Policy.