AgentEng London: The Agent Engineering Conference
Agent Engineering (AgentEng) London is a one-day technical conference dedicated to the engineering behind Agentic AI.
🗓️ Friday 16 October 2026 · 08:00–16:30 ·
📍 Everyman Canary Wharf, London
🎙️ Ten talks · one track · one cinema · limited seats
This is the first conference built around the engineering disciplines behind AI agents: prompt, context, harness, evals, memory, inference, loops, protocols, graphs, search and code.
Agent Engineering
Agent Engineering is emerging as a discipline in its own right. Intelligent AI Agents are getting smarter. Once agents use tools, maintain context, write code, search, remember, interact with other systems and operate over long-running workflows, the model is only one part of the engineering problem. It brings together the engineering practices required to take agents from prototype to reliable production systems:
Prompt Engineering · Context Engineering · Harness Engineering · Eval Engineering · Memory Engineering · Inference Engineering · Loop Engineering · Agentic Engineering · Code Engineering · Protocol Engineering · Graph Engineering · Search Engineering and many more to come.
These disciplines are increasingly interconnected and we will be covering them here at this conference.
💻 Official Website
Explore the Agent Engineering disciplines and detailed conference programme at official website
✨ https://agentengineering.world
🎙️ Confirmed speakers (🇬🇧 🇪🇺 🦾)
The first confirmed speakers include Founders, CEOs, CTOs and engineering leaders from Europe's agent infrastructure companies and more to come.
🎤 Speakers and Talks
Here are the speaker cards with speakers info and talk details.
🎙️ Samuel Colvin (CEO at Pydantic)
Samuel Colvin is Founder & CEO at Pydantic
Talk: Software Development as Constrained Optimization
Abstract:
For most of my career, my job was writing code.Today that's a small part of what I do: AI can take care of the implementation. My job now is to formulate the problem for the AI to solve - choose the metrics to optimize and the constraints to respect - then make it fast and easy for the AI to iterate. This sometimes feels like herding the model: cutting off all the bad routes, so its randomness can only push in the direction of the improvement I want. In this talk, I'll review three real-world projects through this lens, and propose practical ways to set up feedback for the model. This is not a software-factory pitch.
I still write code, and my review is often the most important constraint, but I don't think it makes sense to pretend that me writing and reading all the code is the best way to build software anymore. Luckily, in a world where AIs write the code, the human who formulates the problem is more valuable than the human who wrote the code was a year ago.
🎙️ Andrey Breslav (Founder at CodeSpeak)
Andrey Breslav Founder, CodeSpeak
Talk: If coding is solved, what's left?
Abstract:
A lot of code is generated by agents, and with time, manual coding will go away (almost) completely. What are human engineers going to do then? Will we even have human engineers?
Can agents just do all the planning, coding, reviewing, and testing for us? Do humans have something that LLMs don't?
I think there are some hard limits on what can be delegated, and how. While machines can do a lot of very intelligent work, there are things they won't be doing unless Skynet takes over the world.
🎙️ Ismail Homan Pelaseyed (CTO at Superagent)
Ismail Homan Pelaseyed Co-founder & CTO, Superagent
Talk: Security for AI-Native Developers
Abstract:
AI agents are changing not only how software is built, but how it needs to be secured. This talk explores the emerging security model for AI-native development, where autonomous agents write code, use tools, interact with untrusted data, and increasingly operate across the entire software lifecycle.
🎙️ Alexander Korchagin (Head of Technology Adoption at JetBrains)
Alexander Korchagin - PhD - Head of Technology Adoption at JetBrains
Talk: Agentic Coding in 2026: Watching 300 Engineering Orgs adopting agents
Abstract: Most of what gets presented about agentic coding is what worked or what we hope will work. This talk is the other column.
2026 was the year several comfortable assumptions failed under load. Engineering organisations would converge on a single tooling — they didn't, and the ones that mandated it found their engineers routing around it. Productivity gains would be self-evident — but controlled measurement found experienced developers running slower while reporting they were faster. Betting on one vendor seemed the safe option, in a year with three separate model-access shutoffs. That cost would stay acceptable, right up until it became a top-three adoption barrier.
A field report on what the industry collectively got wrong, what the evidence actually shows, and which of these are now solved versus merely renamed.
🎙️ Tobie Morgan Hitchcock (CEO at SurrealDB)
Tobie Morgan Hitchcock Founder & CEO, SurrealDB
Talk: Memory Is Not a Bigger Context Window: Memory Engineering for Production Agents
Abstract:
Tobie examines why memory engineering, rather than simply expanding the context window, matters for production agents.
🎙️ Jon Bratseth (CEO at Vespa.ai)
Jon Bratseth CEO, Vespa.ai
Talk: The right information at the right time
Abstract:
AI Agents are increasingly bottlenecked on access to the information they need to solve a task. This talk will explain what it takes to enable them to access information that is up to date, complete, accurate, and appropriate for the task at hand.
🎙️Meryem Arik (CEO Doubleword)
Meryem Arik. CEO, Doubleword
Talk: Inference Engineering for Long-Running Agents
Abstract: As models become more capable and agents become more ambitious, long-running agents require abundant, affordable inference. This talk looks at the engineering behind making that possible, from caching and scheduling to throughput and serving efficiency
🎙️Frédéric Barthelet (CTO ALPIC)
Frédéric Barthelet, CTO Alpic
Talk: Browser Use Is the Default. Your UI Is Not Your Agent API.
Abstract: Most harnesses prefer browser use to act on the world, even when an MCP tool exists. This talk shows why that default is bad Agent Experience, how MCP, WebMCP, and CLI give agents a real contract on the same features, and what to ship so callable surfaces win selection over scraping your own product.
🎙️Stas Bichenko (CEO Recombine)
Stas Bichenko CEO Recombine
Talk: Debugging and Optimizing Voice Agents
Abstract: Stas looks at how to identify problems in voice agents and improve their performance.
🎙️Jocelyn D'Arcy (CEO Poindexter Labs)
Jocelyn D'Arcy CEO Poindexter Labs
Talk: Say Less, Grade Harder
Abstract: Jocelyn examines why most agent benchmarks feel artificial, and what changes when tasks are as underspecified as real work.
Checkout Speakers Page on Official website
✨ https://agentengineering.world/speakers
📋 Shape of the Day (Agenda)
Friday 16 October 2026 Everyman, Canary Wharf
08:00–09:00 Doors, registration, tea and pastries
09:00–09:15 Welcome
09:15–09:45 Alexander Korchagin: Agentic Coding in 2026: Watching 300 Engineering Orgs adopting agents
09:50–10:20 Meryem Arik, Doubleword Inference Engineering for Long-Running Agents
10:25–10:55 Samuel Colvin, Pydantic Software Development as Constrained Optimization
11:00–11:15 Networking break
11:15–11:45 Jon Bratseth, Vespa.ai The right information at the right time
11:50–12:20 Andrey Breslav, CodeSpeak If coding is solved, what's left?
12:25–12:55 Stas Bichenko, Recombine Debugging and Optimizing Voice Agents
13:00–14:00 Lunch
14:00–14:30 Frédéric Barthelet, Alpic Browser Use Is the Default. Your UI Is Not Your Agent API.
14:35–15:05 Ismail Pelaseyed, Superagent Security for AI-Native Developers
15:05–15:20 Networking break
15:20–15:50 Jocelyn D’Arcy, Poindexter Labs Say Less, Grade Harder
15:50–16:20 Tobie Morgan Hitchcock, SurrealDB Memory Is Not a Bigger Context Window
16:20–16:30 Closing remarks
16:30 Informal networking at the venue
Schedule may change for speakers travelling from Europe 🇪🇺.
Full programme will be available at: https://agentengineering.world/agenda
🧾 Downloading Your Invoice/Receipt
Once you got your ticket, you can download your invoice and receipt from the Luma payment settings on the Luma web (not on the app)
- Visit https://luma.com/settings/payment
- Find the ticket you paid and click on the three dots
- There you will see an option to download the invoice and receipt for your Agent Engineering ticket.
📍The venue
Everyman
Canary Wharf
Crossrail Place
London E14 5AR
The conference takes place in a boutique cinema rather than a hotel ballroom. Expect a large cinema screen, 7.1 audio and armchair and sofa seating designed for a full day of focused technical talks. Canary Wharf is the financial hub of London served by the Jubilee line, Elizabeth line and DLR. Further information on finding the venue will follow closer to the event.
Breakfast and lunch are included.
P.S The venue is awesome but bit tricky to find so extra guidance will be provided for the attendees and speakers few days before conference so that everyone can find the venue easily.
🙏Code of Conduct
Tickets are not refundable and cannot be transferred without our approval. If you can no longer attend, contact us and we will try to move your place.
💻 Who it is for
This conference is for people who build, operate or lead the development of agent systems. AgentEng London is for AI/ML engineers, AI researchers, founders, CEO/CTO and technical leaders building or operating agent systems.
People who ship agent systems in London and across Europe:
CEO/CTO, Directors and Technical Leaders
AI / ML engineers
Product engineers and technical PMs
Founders and technical leaders
Researchers and applied scientists
If you work on agent infrastructure, developer tools, AI platforms, evaluation, context, memory, inference, protocols, search, coding agents or production agent applications, this is the room.
Registration is curated and subject to approval. Where a registration does not align with that purpose, we may decline it.
🎯 Why Agent Engineering Conference is Different
Single curated track from CEO/CTO and Directors of the Agent Infra Companies. Speaking slots are invited and kept separate from sponsorship. Every talk includes audience Q&A. The room is curated engineers, builders and technical leaders not a mixed expo crowd.
🎛️ FAQ
What is Agent Engineering?
The discipline of designing, building, evaluating and operating AI agents prompts, context, harnesses, evals, memory, inference, loops, protocols, graphs, search and code.How do tickets work?
Apply on this page. Registrations are reviewed to keep the room high-signal, normally within a few hours. You pay only if accepted.Can I book for a team?
Yes. Each person registers. Use Luma invoices after approval for expense.Will talks be recorded?
Check the official site for the current recording policy as the programme is finalised: https://agentengineering.worldIs there networking afterwards?
Yes. Programme ends 16:30; informal networking continues at the venue.Organised with London Agentic AI, the UK Agentic AI community of 5,000+ engineers, and Agent Engineering HQ.
🙌Communities Behind Agent Engineering
🇬🇧About London Agentic AI
London Agentic AI is the UK's original, highly technical community for agent builders, AI engineers, researchers, founders and technical leaders building production-grade agentic systems. Our events are high-signal and implementation-focused. This conference is the founding day that grew out of that series. Sister community Agent Engineering HQ runs technical events in San Francisco on the engineering practices behind agents: prompt, context, harness, memory, evals, inference, loop, protocol, graph, search, and code. Basically *.engineering.
Apply above. Questions: Contact Host, or https://agentengineering.world
#AgentEng #London #AI #AIAgents #AgenticAI