GenAI London 2026 Agenda

GenAI London 2026 Agenda

GenAI London 2026 Agenda

2026 Programme

Hear big ideas
from thought leaders

2026 Programme

Hear big ideas
from thought leaders

Join GenAI London 2026

Apply to speak

10:00 am

11:00 am

Doors Open & Registration

11:00 am

11:15 am

Opening

Opening Remarks

Eric Anderegg, Co-founder at GenAI London
Eric Anderegg

Co-founder

Eric Anderegg, Co-founder of GenAI London, GenAI Zürich and GenAI Berlin, opens the GenAI London 2026 conference at the Queen Elizabeth II Centre, just steps from Big Ben, the Houses of Parliament and Westminster Abbey. His opening remarks set the stage for a day exploring how organisations are putting GenAI into practice.

Opening Remarks

Eric Anderegg, Co-founder at GenAI London
Eric Anderegg

Co-founder

Eric Anderegg, Co-founder of GenAI London, GenAI Zürich and GenAI Berlin, opens the GenAI London 2026 conference at the Queen Elizabeth II Centre, just steps from Big Ben, the Houses of Parliament and Westminster Abbey. His opening remarks set the stage for a day exploring how organisations are putting GenAI into practice.

11:15 am

12:00 pm

The GenAI Frontier: From Capability to Practice

From AI Users to AI Builders: Turning Ideas into Working Tools

Stephanie Anani, Applied AI Architect at OpenAI
Stephanie Anani

Applied AI Architect

AI is expanding who can build software. People who understand a business problem can increasingly turn an idea into a working tool, changing how expertise becomes useful solutions. What does this mean for teams, their skills and the work they choose to do?

In this 10-minute talk, Stephanie Anani follows one illustrative workflow from a recurring business problem to an AI-assisted prototype and a decision about whether it is ready to share. The example explores what domain experts can now do themselves, where engineers and other specialists remain essential, and how to preserve learning and ownership as building becomes easier.

Attendees will leave with three takeaways: how to recognise a small, worthwhile problem to build for; how to distinguish a promising prototype from a tool colleagues can depend on; and how broader building capability can strengthen collaboration and human agency.

From AI Users to AI Builders: Turning Ideas into Working Tools

Stephanie Anani, Applied AI Architect at OpenAI
Stephanie Anani

Applied AI Architect

AI is expanding who can build software. People who understand a business problem can increasingly turn an idea into a working tool, changing how expertise becomes useful solutions. What does this mean for teams, their skills and the work they choose to do?

In this 10-minute talk, Stephanie Anani follows one illustrative workflow from a recurring business problem to an AI-assisted prototype and a decision about whether it is ready to share. The example explores what domain experts can now do themselves, where engineers and other specialists remain essential, and how to preserve learning and ownership as building becomes easier.

Attendees will leave with three takeaways: how to recognise a small, worthwhile problem to build for; how to distinguish a promising prototype from a tool colleagues can depend on; and how broader building capability can strengthen collaboration and human agency.

Answers You Can Trust: Engineering Reliable AI Systems That Earn It

Daniil Shliamov, Principal Product Manager at Perplexity AI
Daniil Shliamov

Principal Product Manager

Anyone can build an AI system that produces plausible answers. Building one people can actually trust is a different engineering discipline. Drawing on lessons from building answer systems at scale, this talk covers the full trust stack.

First, engineering context and orchestrating heterogeneous tools (search, MCP integrations, internal data sources) so answers are grounded in evidence rather than hallucinated. Second, measuring what matters, using groundedness, faithfulness, and tool correctness instead of vanity benchmarks. Third, closing the loop with automated evals, LLM-as-judge scoring, and production feedback pipelines that let the system evaluate itself and propose its own improvements.

Attendees will leave with a practical framework for turning an AI assistant from a confident guesser into a dependable instrument whose answers earn their trust.

Answers You Can Trust: Engineering Reliable AI Systems That Earn It

Daniil Shliamov, Principal Product Manager at Perplexity AI
Daniil Shliamov

Principal Product Manager

Anyone can build an AI system that produces plausible answers. Building one people can actually trust is a different engineering discipline. Drawing on lessons from building answer systems at scale, this talk covers the full trust stack.

First, engineering context and orchestrating heterogeneous tools (search, MCP integrations, internal data sources) so answers are grounded in evidence rather than hallucinated. Second, measuring what matters, using groundedness, faithfulness, and tool correctness instead of vanity benchmarks. Third, closing the loop with automated evals, LLM-as-judge scoring, and production feedback pipelines that let the system evaluate itself and propose its own improvements.

Attendees will leave with a practical framework for turning an AI assistant from a confident guesser into a dependable instrument whose answers earn their trust.

AI Is Only as Good as Its Context: How Organizations Can Become AI-Native

Ankit Jain, Senior Software Engineer at Google DeepMind
Ankit Jain

Senior Software Engineer

Most organizations already have AI tools, but few work in an AI-native way. In this talk I'll go through what I think it takes to get there. It starts with people. They have to give up ways of working they're comfortable with. Power users need room to build tools and show others how it's done. And every few months we need to check again what AI can reliably do. I think the real bottleneck, though, is organizational context. An agent can only see the context we give it. Most of what an org knows is never written down, is stuck in silos, or loses the framing that gives it meaning. I'll cover why this is hard and what the solution is starting to look like.

AI Is Only as Good as Its Context: How Organizations Can Become AI-Native

Ankit Jain, Senior Software Engineer at Google DeepMind
Ankit Jain

Senior Software Engineer

Most organizations already have AI tools, but few work in an AI-native way. In this talk I'll go through what I think it takes to get there. It starts with people. They have to give up ways of working they're comfortable with. Power users need room to build tools and show others how it's done. And every few months we need to check again what AI can reliably do. I think the real bottleneck, though, is organizational context. An agent can only see the context we give it. Most of what an org knows is never written down, is stuck in silos, or loses the framing that gives it meaning. I'll cover why this is hard and what the solution is starting to look like.

12:00 pm

12:30 pm

Panel Discussion

From GenAI Tools to GenAI-Native Organisations

Ankit Jain, Senior Software Engineer at Google DeepMind
Ankit Jain

Senior Software Engineer

Denis Samuylov, Co-Founder at GenAI London
Denis Samuylov

Co-Founder

Stephanie Anani, Applied AI Architect at OpenAI
Stephanie Anani

Applied AI Architect

From GenAI Tools to GenAI-Native Organisations

Ankit Jain, Senior Software Engineer at Google DeepMind
Ankit Jain

Senior Software Engineer

Denis Samuylov, Co-Founder at GenAI London
Denis Samuylov

Co-Founder

Stephanie Anani, Applied AI Architect at OpenAI
Stephanie Anani

Applied AI Architect

Stay tuned – more will be announced here shortly.

12:45 pm

01:10 pm

GenAI in Marketing

AI and the battle for creativity

Ned Seagrim, AI Lead, UK&I at Ogilvy
Ned Seagrim

AI Lead, UK&I

Every industry is asking the same question: what can we do that AI can’t? For the marketing and creative industries, solace has been taken in the belief that the answer lies in the sacred human skill of creativity. But what if AI can in fact be creative? What would that mean for how we use the technology and do marketing in the future? Would this be existential for the future of creative agencies or would this unlock a new approach to marketing?

AI and the battle for creativity

Ned Seagrim, AI Lead, UK&I at Ogilvy
Ned Seagrim

AI Lead, UK&I

Every industry is asking the same question: what can we do that AI can’t? For the marketing and creative industries, solace has been taken in the belief that the answer lies in the sacred human skill of creativity. But what if AI can in fact be creative? What would that mean for how we use the technology and do marketing in the future? Would this be existential for the future of creative agencies or would this unlock a new approach to marketing?

99 problems, and AI solved the wrong one.

Naji El-Arifi, Director, Agentic Strategy & Design at WPP
Naji El-Arifi

Director, Agentic Strategy & Design

AI has made doing things almost free. You can brief a tool, walk away and come back an hour later to a polished answer. The catch is that it might be the answer to the wrong question. In marketing, that means a team heading confidently in the wrong direction, at AI speed. This talk is about the one habit that matters more now than before AI: being sure you know what problem you’re solving before you start building.

99 problems, and AI solved the wrong one.

Naji El-Arifi, Director, Agentic Strategy & Design at WPP
Naji El-Arifi

Director, Agentic Strategy & Design

AI has made doing things almost free. You can brief a tool, walk away and come back an hour later to a polished answer. The catch is that it might be the answer to the wrong question. In marketing, that means a team heading confidently in the wrong direction, at AI speed. This talk is about the one habit that matters more now than before AI: being sure you know what problem you’re solving before you start building.

01:10 pm

01:45 pm

Panel Discussion

Marketing Transformation: People, platforms and perseverance.

Dr Laura Weis, Global Head of Human–AI Strategy & Innovation at WPP
Dr Laura Weis

Global Head of Human–AI Strategy & Innovation

Kate Mackie, Chief Marketing Strategy & Operations Officer at EY
Kate Mackie

Chief Marketing Strategy & Operations Officer

Mark Dewings, CEO at Ogilvy Consulting UK
Mark Dewings
Ozgur Kirazci, Sr Marketing Director, EMEA Platforms & Devices; UKI Devices & Services at Google
Ozgur Kirazci

Sr Marketing Director, EMEA Platforms & Devices; UKI Devices & Services

Mark Dewings (CEO, Ogilvy Consulting UK) will host Kate Mackie (Chief Marketing Strategy & Operations Officer, EY), Özgür Kirazci (Senior Marketing Director, EMEA Platforms and Devices, Google) and Dr Laura Weiss (WPP, Human-AI Strategy & Workforce Planning Lead) to hear their views on AI-enabled marketing transformation. The discussion will cover what has actually changed, how organisations should transform, views on the future marketing organisation, and more.

Marketing Transformation: People, platforms and perseverance.

Dr Laura Weis, Global Head of Human–AI Strategy & Innovation at WPP
Dr Laura Weis

Global Head of Human–AI Strategy & Innovation

Kate Mackie, Chief Marketing Strategy & Operations Officer at EY
Kate Mackie

Chief Marketing Strategy & Operations Officer

Mark Dewings, CEO at Ogilvy Consulting UK
Mark Dewings
Ozgur Kirazci, Sr Marketing Director, EMEA Platforms & Devices; UKI Devices & Services at Google
Ozgur Kirazci

Sr Marketing Director, EMEA Platforms & Devices; UKI Devices & Services

Mark Dewings (CEO, Ogilvy Consulting UK) will host Kate Mackie (Chief Marketing Strategy & Operations Officer, EY), Özgür Kirazci (Senior Marketing Director, EMEA Platforms and Devices, Google) and Dr Laura Weiss (WPP, Human-AI Strategy & Workforce Planning Lead) to hear their views on AI-enabled marketing transformation. The discussion will cover what has actually changed, how organisations should transform, views on the future marketing organisation, and more.

02:00 pm

02:25 pm

GenAI Tech Stack

The Database Estate Nobody Provisioned

James Tsao, Sr. Staff Sales Engineer at Cockroach Labs
James Tsao

Sr. Staff Sales Engineer

AI agents are beginning to create and remove databases programmatically, while many teams still rely on processes built for people to provision and manage them. As agent traffic grows, some infrastructure will absorb the demand; other teams may face rising costs and operational sprawl.

This session explores what it takes to manage one database estate across workloads created by both people and agents, and asks a practical question: what does your database estate cost today, and what might it cost at ten times the agent traffic?

The Database Estate Nobody Provisioned

James Tsao, Sr. Staff Sales Engineer at Cockroach Labs
James Tsao

Sr. Staff Sales Engineer

AI agents are beginning to create and remove databases programmatically, while many teams still rely on processes built for people to provision and manage them. As agent traffic grows, some infrastructure will absorb the demand; other teams may face rising costs and operational sprawl.

This session explores what it takes to manage one database estate across workloads created by both people and agents, and asks a practical question: what does your database estate cost today, and what might it cost at ten times the agent traffic?

The Art of Evals: Building a Benchmark From Zero

Nick Mecklenburg, Co-Founder & CTO at Whitney AI
Nick Mecklenburg

Co-Founder & CTO

How do you evaluate an AI system when no suitable benchmark exists and the data you need is difficult to obtain?

This talk follows the construction of a semi-synthetic benchmark for detecting errors in US medical bills. Starting with public healthcare data, we reconstructed clinical encounters, incorporated hospital pricing, and introduced controlled billing errors. Each step required decisions about realism, coverage, and what “good performance” should mean.

Metric design proved just as important as the dataset; catching ten small billing errors while missing one expensive error can reward a model without helping the patient. We designed metrics to better capture patient financial utility, examined performance across models, and identified limitations that still require expert validation.

Takeaways:

  • Build evaluation data through grounded synthesis, when real examples are not available.

  • Assess benchmarks for coverage, alignment, realism, and room for improvement.

  • Choose metrics that reflect user outcomes, including the consequences of false positives.

  • Recognize the pros and cons of synthetic eval data.

The Art of Evals: Building a Benchmark From Zero

Nick Mecklenburg, Co-Founder & CTO at Whitney AI
Nick Mecklenburg

Co-Founder & CTO

How do you evaluate an AI system when no suitable benchmark exists and the data you need is difficult to obtain?

This talk follows the construction of a semi-synthetic benchmark for detecting errors in US medical bills. Starting with public healthcare data, we reconstructed clinical encounters, incorporated hospital pricing, and introduced controlled billing errors. Each step required decisions about realism, coverage, and what “good performance” should mean.

Metric design proved just as important as the dataset; catching ten small billing errors while missing one expensive error can reward a model without helping the patient. We designed metrics to better capture patient financial utility, examined performance across models, and identified limitations that still require expert validation.

Takeaways:

  • Build evaluation data through grounded synthesis, when real examples are not available.

  • Assess benchmarks for coverage, alignment, realism, and room for improvement.

  • Choose metrics that reflect user outcomes, including the consequences of false positives.

  • Recognize the pros and cons of synthetic eval data.

02:35 pm

03:00 pm

Networking Break

03:00 pm

03:15 pm

Inspirational Talk

AI to Real and Back to AI

Guli Silberstein, Artist-Founder at Guli Silberstein Art
Guli Silberstein

Artist-Founder

The Neural Forest digital art project introduces a pioneering "phygital" approach that materialises digital concepts into tangible, premium video sculptures. While advanced generative AI tools expand the limits of imagination, their visual art outputs typically remain trapped as fleeting pixels on a screen. This project bridges that virtual-physical divide, transforming fluid digital files into high-value, physical art assets.

The production pipeline seamlessly moves from initial concept to digital fabrication. In a collaboration between human and machine, 3D mesh modelling is used, continuing through industrial 3D printing, and high-gloss chrome plating. And raw organic structures are transformed into reflective chrome-finished sculptures. These physical forms securely cradle LCD screens running custom video loop artworks made from a unique combination of photography, AI, and datamoshing.

Looking ahead, Neural Forest heralds the transition towards a metamorphic engine that bridges digital concepts and tangible reality. It points to a future of pure creation, where machines will materialise ideas into diverse physical forms, ranging from synthetic compounds, and eventually, the human form itself. This presentation demonstrates a sophisticated pipeline where AI and nature converge, offering an exclusive look at the future of high-end, tangible digital art.

AI to Real and Back to AI

Guli Silberstein, Artist-Founder at Guli Silberstein Art
Guli Silberstein

Artist-Founder

The Neural Forest digital art project introduces a pioneering "phygital" approach that materialises digital concepts into tangible, premium video sculptures. While advanced generative AI tools expand the limits of imagination, their visual art outputs typically remain trapped as fleeting pixels on a screen. This project bridges that virtual-physical divide, transforming fluid digital files into high-value, physical art assets.

The production pipeline seamlessly moves from initial concept to digital fabrication. In a collaboration between human and machine, 3D mesh modelling is used, continuing through industrial 3D printing, and high-gloss chrome plating. And raw organic structures are transformed into reflective chrome-finished sculptures. These physical forms securely cradle LCD screens running custom video loop artworks made from a unique combination of photography, AI, and datamoshing.

Looking ahead, Neural Forest heralds the transition towards a metamorphic engine that bridges digital concepts and tangible reality. It points to a future of pure creation, where machines will materialise ideas into diverse physical forms, ranging from synthetic compounds, and eventually, the human form itself. This presentation demonstrates a sophisticated pipeline where AI and nature converge, offering an exclusive look at the future of high-end, tangible digital art.

03:15 pm

04:00 pm

GenAI in the Enterprise

Institutional Memory: Turning AI Coding Assistants into a Team's Collective Brain

Filipe Albero Pomar, Engineering Manager at Maersk
Filipe Albero Pomar

Engineering Manager

Every engineer using Claude Code or Copilot generates a session full of hard-won context: gotchas, conventions, decisions. Today, that knowledge evaporates the moment the session ends, and the next engineer rediscovers it the hard way.

This talk shares an early but promising system for capturing that context and turning it into shared team memory. Sessions are logged centrally, then processed into durable, tagged knowledge fed back into future sessions.

Filipe will cover what's working, what's still limited, and the harder problems ahead: keeping extracted knowledge accurate rather than confidently wrong, who owns retiring stale entries, and stopping shared memory becoming another thing to maintain. A practitioner's account from an engineering manager building this now, not a finished product pitch.

Key takeaways

  • A working pattern for capturing team knowledge from AI coding sessions

  • Why "most frequent" metrics hide the moments that matter most

  • The real bottleneck: extraction quality, not storage

More about Filipe: https://alpomar.dev

Institutional Memory: Turning AI Coding Assistants into a Team's Collective Brain

Filipe Albero Pomar, Engineering Manager at Maersk
Filipe Albero Pomar

Engineering Manager

Every engineer using Claude Code or Copilot generates a session full of hard-won context: gotchas, conventions, decisions. Today, that knowledge evaporates the moment the session ends, and the next engineer rediscovers it the hard way.

This talk shares an early but promising system for capturing that context and turning it into shared team memory. Sessions are logged centrally, then processed into durable, tagged knowledge fed back into future sessions.

Filipe will cover what's working, what's still limited, and the harder problems ahead: keeping extracted knowledge accurate rather than confidently wrong, who owns retiring stale entries, and stopping shared memory becoming another thing to maintain. A practitioner's account from an engineering manager building this now, not a finished product pitch.

Key takeaways

  • A working pattern for capturing team knowledge from AI coding sessions

  • Why "most frequent" metrics hide the moments that matter most

  • The real bottleneck: extraction quality, not storage

More about Filipe: https://alpomar.dev

Engineering Leaders Must Build Again: Why, and How to Start

Ronen Elman, Senior Leader and Advisor at Independent
Ronen Elman

Senior Leader and Advisor

Agentic AI is changing the fundamentals of engineering, and engineering leaders who haven’t stayed close enough to the craft run the risk of leading their teams through a change they don’t fully understand. Spending some time building again is the best way to get exposure to this new, evolving engineering practice, and so be a better leader for your team. But that can feel like a daunting task.

This talk is about how I got back to building and experimenting with agentic flows after years away from hands-on work. I’ll share the blueprint I followed, and the specific tools and resources I used along the way.

You’ll leave with a clear reason to start, and a way to do it starting tomorrow.

Engineering Leaders Must Build Again: Why, and How to Start

Ronen Elman, Senior Leader and Advisor at Independent
Ronen Elman

Senior Leader and Advisor

Agentic AI is changing the fundamentals of engineering, and engineering leaders who haven’t stayed close enough to the craft run the risk of leading their teams through a change they don’t fully understand. Spending some time building again is the best way to get exposure to this new, evolving engineering practice, and so be a better leader for your team. But that can feel like a daunting task.

This talk is about how I got back to building and experimenting with agentic flows after years away from hands-on work. I’ll share the blueprint I followed, and the specific tools and resources I used along the way.

You’ll leave with a clear reason to start, and a way to do it starting tomorrow.

Proving the Decision: What Agentic AI Requires of Human Judgement Now

Dr Joanna Michalska, Founder & Managing Director at Ethica Group Ltd
Dr Joanna Michalska

Founder & Managing Director

Agentic AI moves the human decision earlier and compresses the time to make it, raising the bar for what counts as judgement rather than reflex. The question now is whether the person still in the loop can see clearly, hold their position under speed and pressure, and exercise authority they can prove. This session sets out what agentic AI requires of human judgement now: decisions tested under time compression, evidenced when they matter, and owned by a named person who can show they held. Humans still decide. The proof required to show that judgement holds is what has changed. Key takeaways:

  • Judgement counts as authority only when it has been tested under real pressure and time compression.

  • The human role in agentic systems moves earlier: judgement now happens before execution.

  • What has changed is the proof required: judgement must be shown to hold, not assumed because a human was present.

Proving the Decision: What Agentic AI Requires of Human Judgement Now

Dr Joanna Michalska, Founder & Managing Director at Ethica Group Ltd
Dr Joanna Michalska

Founder & Managing Director

Agentic AI moves the human decision earlier and compresses the time to make it, raising the bar for what counts as judgement rather than reflex. The question now is whether the person still in the loop can see clearly, hold their position under speed and pressure, and exercise authority they can prove. This session sets out what agentic AI requires of human judgement now: decisions tested under time compression, evidenced when they matter, and owned by a named person who can show they held. Humans still decide. The proof required to show that judgement holds is what has changed. Key takeaways:

  • Judgement counts as authority only when it has been tested under real pressure and time compression.

  • The human role in agentic systems moves earlier: judgement now happens before execution.

  • What has changed is the proof required: judgement must be shown to hold, not assumed because a human was present.

04:15 pm

04:50 pm

The Next Wave of European GenAI

What Makes AI Deployment Work vs. Not

Husayn Kassai, CEO and Founder at Ollo (Askollo Ltd)
Husayn Kassai

CEO and Founder

Everyone is building with AI, but few are deploying it successfully. Join Husayn Kassai, Founder & CEO of Ollo, for a practical look at why so many AI initiatives stall: unclear decision-making, weak ownership, and bots bolted onto workflows that are already full of friction.

Husayn introduces the AI "harness": the organisational layer of judgment, governance, context and tooling that turns what models can do into real business impact. You'll learn how to define what good looks like, who should own AI-assisted decisions, and how to give systems the access and context they need.

Attendees will leave with a framework for choosing between generalist tools, specialist products and custom workflows. They'll also see why the ability to deploy AI must be built in-house. Successful AI isn't model-led; it's organisation-led.

What Makes AI Deployment Work vs. Not

Husayn Kassai, CEO and Founder at Ollo (Askollo Ltd)
Husayn Kassai

CEO and Founder

Everyone is building with AI, but few are deploying it successfully. Join Husayn Kassai, Founder & CEO of Ollo, for a practical look at why so many AI initiatives stall: unclear decision-making, weak ownership, and bots bolted onto workflows that are already full of friction.

Husayn introduces the AI "harness": the organisational layer of judgment, governance, context and tooling that turns what models can do into real business impact. You'll learn how to define what good looks like, who should own AI-assisted decisions, and how to give systems the access and context they need.

Attendees will leave with a framework for choosing between generalist tools, specialist products and custom workflows. They'll also see why the ability to deploy AI must be built in-house. Successful AI isn't model-led; it's organisation-led.

AI Evals at 80M Users a Month: Scaling Expert Judgement with LLM Judges

Alexander Taboriskiy, CEO & Founder at Mentiora
Alexander Taboriskiy

CEO & Founder

Vladislav Nedosekin, Director of Engineering, Intelligence Systems at Flo Health
Vladislav Nedosekin

Director of Engineering, Intelligence Systems

Evaluating the customer-facing AI for a women's health app used by 80 million people a month means no team of experts can read every answer. So LLM judges do the reading, and that only works if the judges decide the way the experts would. Out of the box, they often don't: a judge with a sensible rubric can agree with experts barely better than chance.

Health adds a second problem. A judge that is too strict blocks answers people need, and one that is too lenient lets unsafe ones through. Flo and Mentiora split the work in two. Medical judges, built with clinicians, decide whether an answer is safe. Usefulness judges, built with product experts, decide whether it actually helps. The two pull in opposite directions. When we tuned a model to pass the medical judges, it got safer, but less useful. Together the judges form the foundation for a quality loop that raises the customer experience.

In this talk we show how we got from a small sample of expert labels to judges that hold up against those experts. We cover how to measure a judge's agreement with experts before you trust it, and how to search across judge and aggregation combinations instead of hand-tuning one prompt. We show how to combine several narrow judges into one release verdict with hard gates, and why safety and usefulness need separate verdicts rather than one blended score. We finish with where humans still have to stay in the loop.

AI Evals at 80M Users a Month: Scaling Expert Judgement with LLM Judges

Alexander Taboriskiy, CEO & Founder at Mentiora
Alexander Taboriskiy

CEO & Founder

Vladislav Nedosekin, Director of Engineering, Intelligence Systems at Flo Health
Vladislav Nedosekin

Director of Engineering, Intelligence Systems

Evaluating the customer-facing AI for a women's health app used by 80 million people a month means no team of experts can read every answer. So LLM judges do the reading, and that only works if the judges decide the way the experts would. Out of the box, they often don't: a judge with a sensible rubric can agree with experts barely better than chance.

Health adds a second problem. A judge that is too strict blocks answers people need, and one that is too lenient lets unsafe ones through. Flo and Mentiora split the work in two. Medical judges, built with clinicians, decide whether an answer is safe. Usefulness judges, built with product experts, decide whether it actually helps. The two pull in opposite directions. When we tuned a model to pass the medical judges, it got safer, but less useful. Together the judges form the foundation for a quality loop that raises the customer experience.

In this talk we show how we got from a small sample of expert labels to judges that hold up against those experts. We cover how to measure a judge's agreement with experts before you trust it, and how to search across judge and aggregation combinations instead of hand-tuning one prompt. We show how to combine several narrow judges into one release verdict with hard gates, and why safety and usefulness need separate verdicts rather than one blended score. We finish with where humans still have to stay in the loop.

Brand craft, not slop: turning AI generation into usable business content

Fraser Bell, CEO at Wondercraft
Fraser Bell

Generative AI can produce a video or voiceover in seconds. Producing content your brand can actually use is a different job. Ask AI for the same thing twice and you’ll get two different results: a new read of the script, a new pronunciation of your product name, a different look for your product. That’s slop.

Fraser Bell, CEO of Wondercraft, breaks down what it takes to go from AI generation to consistent, brand-ready video and audio. He’ll cover the craft decisions that make the difference, from voice selection and custom pronunciations to editing control, and where AI should do the work versus where a person should.

He’ll also explain why finishing a piece takes more than one model, and the legal questions brands should ask before publishing AI content.

Key takeaways:

  • Practical advice for making AI audio and video brand-ready

  • Which steps to automate and which still need human judgement

  • How to set up tools and workflow so a team can finish the job

  • The legal protections to have in place before AI content goes out

Brand craft, not slop: turning AI generation into usable business content

Fraser Bell, CEO at Wondercraft
Fraser Bell

Generative AI can produce a video or voiceover in seconds. Producing content your brand can actually use is a different job. Ask AI for the same thing twice and you’ll get two different results: a new read of the script, a new pronunciation of your product name, a different look for your product. That’s slop.

Fraser Bell, CEO of Wondercraft, breaks down what it takes to go from AI generation to consistent, brand-ready video and audio. He’ll cover the craft decisions that make the difference, from voice selection and custom pronunciations to editing control, and where AI should do the work versus where a person should.

He’ll also explain why finishing a piece takes more than one model, and the legal questions brands should ask before publishing AI content.

Key takeaways:

  • Practical advice for making AI audio and video brand-ready

  • Which steps to automate and which still need human judgement

  • How to set up tools and workflow so a team can finish the job

  • The legal protections to have in place before AI content goes out

04:50 pm

05:15 pm

Panel Discussion

Can Europe Lead the Future of GenAI?

Alexander Taboriskiy, CEO & Founder at Mentiora
Alexander Taboriskiy

CEO & Founder

Eric Anderegg, Co-founder at GenAI London
Eric Anderegg

Co-founder

Fraser Bell, CEO at Wondercraft
Fraser Bell
Husayn Kassai, CEO and Founder at Ollo (Askollo Ltd)
Husayn Kassai

CEO and Founder

Can Europe Lead the Future of GenAI?

Alexander Taboriskiy, CEO & Founder at Mentiora
Alexander Taboriskiy

CEO & Founder

Eric Anderegg, Co-founder at GenAI London
Eric Anderegg

Co-founder

Fraser Bell, CEO at Wondercraft
Fraser Bell
Husayn Kassai, CEO and Founder at Ollo (Askollo Ltd)
Husayn Kassai

CEO and Founder

05:20 pm

05:50 pm

Award Ceremony

05:50 pm

06:00 pm

Closing

Closing Remarks

Denis Samuylov, Co-Founder at GenAI London
Denis Samuylov

Co-Founder

Eric Anderegg, Co-founder at GenAI London
Eric Anderegg

Co-founder

The closing ceremony of GenAI London 2026.

Closing Remarks

Denis Samuylov, Co-Founder at GenAI London
Denis Samuylov

Co-Founder

Eric Anderegg, Co-founder at GenAI London
Eric Anderegg

Co-founder

The closing ceremony of GenAI London 2026.

Sponsors and partners

Supported by forward-thinking companies shaping the GenAI transformation

Our esteemed partners are pushing the boundaries of what's possible with Generative AI. Be part of the GenAI leaders shaping what’s next.

Become a partner

GenAI Masters

Cockroach Labs logo
Skyflow logo

GenAI Next

Bruin logo

Special Partner for Ethical AI, Cybersecurity Governance & Human Rights

ICT4peace Foundation logo

Sponsors and partners

Supported by forward-thinking companies shaping the GenAI transformation

Our esteemed partners are pushing the boundaries of what's possible with Generative AI. Be part of the GenAI leaders shaping what’s next.

Become a partner

Special Partner

ICT4peace Foundation logo

Community Partner

SheCanCode logo
AI for the Rest of Us logo
Women Coding Community logo
She Bytes Back logo
Community Stack logo
The AI Exchange (an AMP Community) logo
London DevOps logo
WeBloom logo
Seotistics logo
GenAI Zürich logo
Salama Belghali Consulting logo
Switzerland Global Enterprise logo

Join us

Europe’s go-to conference for GenAI leaders and enthusiasts

Attend GenAI London to stay at the forefront of Generative AI, connect with the minds shaping the technology’s future, and explore its real-world impact across industries.

Join GenAI London 2026

Become a partner

Apply to speak

GENAI

London

2026

Join GenAI London 2026

Become a partner

Apply to speak

GENAI LONDON '26 WARM UP

Join us

Europe’s go-to conference for GenAI leaders and enthusiasts

Attend GenAI London to stay at the forefront of Generative AI, connect with the minds shaping the technology’s future, and explore its real-world impact across industries.

Join GenAI London 2026

Become a partner

Apply to speak

GENAI

London

2026

Join GenAI London 2026

Become a partner

Apply to speak

GENAI LONDON '26 WARM UP

Join us

Europe’s go-to conference for GenAI leaders and enthusiasts

Attend GenAI London to stay at the forefront of Generative AI, connect with the minds shaping the technology’s future, and explore its real-world impact across industries.

Join GenAI London 2026

Become a partner

Apply to speak

GENAI

London

2026

Join GenAI London 2026

Become a partner

Apply to speak

GENAI LONDON '26 WARM UP