.png%3Ffit%3Dcover&w=3840&q=75)
.png%3Ffit%3Dcover&w=3840&q=75)
MACH X: Amsterdam
Picnic was built as a software-native retailer, with technology spanning the full journey from farm to fork. MACH principles made that stack modular, scalable, and fast-moving, but AI agents change the equation: software no longer just exposes capabilities through APIs, it lets agents discover, orchestrate, and evolve them. Daniel Gebler, CTO of Picnic, shares what that shift has meant for Picnic's architecture in practice, and offers a framework attendees can use to assess where their own stack is ready for agentic orchestration, and where it isn't.
As JD Sports transforms the spine of its digital stack across every brand and region, the business is embracing composability. The driver is the customer: young, mobile-first, and quick to test whatever is new, from AI assistants to whatever discovery looks like next. Serving that customer means the flexibility to prioritize agentic services, new discovery tools, automated and enriched content, and optimization across surfaces we do not yet understand. Antonia will walk through how the stack is being rethought so JD can move at the speed its customers are accelerating towards.
Virgin media O2 has modernized its customer-facing digital estate by introducing composable capabilities across web, mobile and contact-centre platforms while keeping critical legacy systems unchanged. This keynote explore how this pragmatic approach created the foundation for agentic services, supporting customers, business teams and technology colleagues as the CIO organisation embarks on its next phase of transformation.
As adoption accelerates, it's time to move from reacting to proactively building a new operating model - How does a business create a new profitable model among a sea of constantly changing technology which changes performance and price almost daily? Sam's keynote and workshop will help delegates plan for the short and medium term.
In an increasingly rapid transformation of SaaS, tooling and commercials, the relationships between client and vendor will transform equally quickly. We start our track by asking those involved the key questions about accountability, hallucinations, outcomes, infrastructure and security.
Agentic AI, Tested on Real Travel Problems: Results from the easyJet Holidays & Your Golf Travel Hackathon
For eight weeks, the MACH Alliance community has been building alongside easyJet Holidays and Your Golf Travel, solving for real problems: unifying review data, wiring an interactive points-of-interest map, triaging customer email with a human in the loop, and building natural language search for trip planning.
This session is the first look at what they built: agentic AI tested against real data and real constraints. You'll see what worked, what didn't, and leave with patterns you can take back to your own team.
Germany’s beauty powerhouse continues to conquer Europe’s online beauty market, powered by a lean and agile technology stack launched in 2021. With a composable core, Flaconi has expanded internationally, grown revenue & margin significantly, and maintained a superb NPS score throughout. Now, with agents in play, the business has plans for further expansion, while focusing on being lean in development, rigorous in experimentation, and relentless in the pursuit of efficient technology use.
One of the first key areas of AI impacting the business is in Business intelligence. The business wants "Claude" self-service BI, but this comes at some increased risk - albeit some lowered costs. We'll talk about commercial and technical risk, upside and where "control" lies in this new paradigm.
Most teams know agentic AI is coming but haven't actually built with it yet, which makes it hard to separate the real patterns from the hype. This lab opens with the four core agent architecture patterns, then moves straight into hands-on building with them. Open to business users and technical builders alike, attendees leave having built a working agent, not just watched a slide about one.
EasyJet are deep in it, understanding how best to harness yet govern the myriad AI programs that are underway. Every element needs rethinking, from budgeting and partner contracts to model usage and "agent-to-agent" moments. In a truly Chatham House session, the EasyJet team will share where they have got to, and then we encourage a full and frank debate about where next.
Emilija will offer an unfiltered walk-through of a total digital stack re-platform, with multiple "pivots" and all the genuine complexity of region, language, SKU and connectivity. She'll touch on why AI adoption doesn’t happen just because the CTO says so, and the joy of changing e-commerce vendor six months into a project. There were bumps (some closer to sinkholes) but it mostly worked out in the end. Come for the keynote - stay for the therapy session.
Both Solventum and Dr. Martens have retained a monolithic platform in SAP, but decomposed elements and surrounded the core with AI components. These are stories that are "real life" for many brands, perhaps living with inherited core systems but ambitious to accelerate their digital journey and harness the new tools.
Security teams are now adopting responsibility for helping the business to navigate new potential sources of threat, from unreliable data entering core systems, through to agents acting autonomously with potentially damaging consequences. Our panelists will debate:
- Core process changes needed in order to enable (rather than disable) the business
- Creating new foundations for threats that are unpredictable
- Educating the business leadership (advising vs. prescribing)
- Preparing for Agentic transactions & the new world
Simon Edwards
Cybersecurity Architect at Charles Tyrwhitt
For IFM, a 57-year-old B2B hardware manufacturer, AI-powered selling didn't start with a chatbot, it started with data housekeeping. The team will show how turning 2.5 million scattered assets into a clean, automated content supply chain has become the unglamorous foundation for Smart Search, IFM's AI recommendation tool that now closes knowledge gaps its own product data can't. The takeaway: for B2B leaders chasing AI in sales, the real work is upstream.
The models are now embedded in every engineering team, allowing for scale, speed and ambition for "owned" tech previously unheard of. Build to replace the SaaS? Upskill product teams to build agents that last? Our leaders from across Engineering, Product and General technology will take on this tricky and fast-moving subject
PhotoSi have a newly minted approach towards customer incentives, created alongside partners their view is that loyalty has no place in digital, but targeted "agentic" incentives will yield better long term outcomes and will avoid brand damage through over-discounting.
An intense panel session focused on the core building blocks required to be successful in new discovery landscape: Product data, Owned Content, and an understanding of the context of the landscape. The panelists will discuss social context, semantics and the new search models, and how brands can own their own contextual layer.
12 tables, 6 Subjects:
The role of architects is fundamentally changing. Underlying technologies might remain relatively unchanged for a while, but the avalanche of new tools and new data flows needs a constant level of oversight & planning. Our architects discussion will work through the challenges, and establish ways in which we can all deal with the influx of new native AI.
As AI redefines discovery in digital, so the roles that are needed to optimize content, campaigns and creative need re-thinking. Josh will construct an argument for a new role - the Marketing Engineer - with the skills and autonomy to execute and win the battle for new customers
