Built for AI at Scale: Real-Time Systems Without Limits

Join us for an evening focused on the craft of building modern infrastructure and data systems that scale with speed and reliability. Through real-life case studies and technical deep dives, engineering leaders from *Aerospike*, *Lemonade*, and *Bigabid* will share how their teams approach platform engineering, API modernization, and real-time data performance at scale. Come learn from the experiences of industry experts, exchange ideas with peers, and gain practical insights you can apply to your own systems and architecture challenges. Agenda: 17:30 – 18:00 - Welcome & networking 18:00-18:20 - What we learnt modernising our database API and how we’re planning on avoiding the same mistakes again Tim Faulkes, Chief Developer Advocate, Aerospike When you have a public-facing API, backwards compatibility is mandatory. But technical debt accumulates quickly in this scenario and after 15 years of technical debt, we decided to create a whole new API. Join us to hear the lessons we learnt on architecting an API for readability, maintainability and scalability. We will discuss the challenges of the original API and how we overcame them in the new API (and hopefully made them more future-proof). 18:20-18:50 - Infrastructure as a Product: Platform Engineering at Lemonade Jasmine Nouriel - Staff Platform Engineer @ Lemonade What happens when you treat infrastructure not as plumbing, but as a product? At Lemonade, we’ve taken Infrastructure as Code and self-service to the extreme, building a platform that empowers every developer to own their infrastructure. Join us to discover how a team of just five engineers scaled Platform Engineering to serve hundreds of developers and thousands of deployments, using Pulumi and product-thinking to turn infrastructure into a delightful, autonomous experience. 18:50-19:20 - Serving Smarter: Re-Architecting for Sub-Millisecond Feature Access Ido Nadler, Big Data Team Lead, Bigabid Daniel Miodownik, Director Of Data Infrastructure, Bigabid Behind every real-time decision lies a battle with data latency. In this session, we’ll walk through how Bigabid’s engineering team re-architected its Serving Layer to deliver sub-millisecond feature access for machine learning models at scale. We’ll dive into how we tackled cold-start issues, optimized high-speed lookups, and built a system that scales effortlessly with growth. Expect an inside look at the architectural choices, performance trade-offs, and real-world results that turned our Serving Layer into a low-latency, cost-effective engine powering real-time intelligence 19:20 – 20:00 - Networking