Agents in Production: Scaling Enterprise AI (SurrealDB x NVIDIA @ LTW)
Agents in Production: Scaling Enterprise AI (SurrealDB x NVIDIA @ LTW)
London Tech Week is full of conversations about what enterprise AI could do. This is the event for the engineers who have to build it.
Specifically: agents that hold up past the demo. The ones that need memory that persists across sessions, context that doesn't collapse under load, behaviour that's auditable when something goes wrong, and infrastructure that can actually take the weight. That's a harder problem than most will admit - and we're going to get into it properly.
Two engineering teams, one evening. Bring your hardest questions. Pizza and drinks on us.
Talks
Tobie Morgan Hitchcock (Co-Founder & CEO, SurrealDB)
Knowledge, Memory, Cognition: The Journey Towards Agent Consciousness
Every agent demo has a memory. Most of them fall apart the moment a second session starts, the load goes up, or someone asks why it said what it said. The usual fix is to bolt a vector store onto a graph onto an event log and trust the middleware to keep them honest - which isn't memory, it's plumbing, and it drifts.
Tobie makes the case for a different approach: model memory the way a mind actually does it - knowledge, memory, cognition - and put the whole thing inside one transactional database. Provenance on every fact, supersession instead of overwrites, an audit trail you can query, and memory that keeps working between conversations. One ACID transaction per turn, from a laptop to a cluster.
No magic, and no claims about consciousness - just the substrate an agent needs to remember, and why it belongs in the database rather than around it.
Martin Schaer (Solutions Engineer, SurrealDB)
Context layer and agentic memory in SurrealDB
In the rapidly evolving field of AI, retrieval is only half the battle. The other half? Knowing how to move through your data - deterministically, intelligently, and at scale.
This is a follow-up to our popular "How to Build a Knowledge Graph for AI" blog post. This time, we go deeper: instead of building the graph, we focus on navigating it - using an LLM to generate queries that traverse a rich, structured knowledge graph and return grounded, deterministic answers.
Ziv Ilan (Super AI team, NVIDIA)
Agents You Can Actually Trust: Running Always-On Autonomous Agents Locally with NemoClaw, OpenShell & Nemotron
Most agent demos break the moment they leave the laptop. This talk shows how a fully open-source stack — NVIDIA's OpenShell runtime, the NemoClaw reference stack, and open Nemotron 3 models — lets you run autonomous agents that are secure-by-design, observable, and production-ready. We'll cover the architecture, what breaks at scale, and why building on open models and an open runtime keeps you in control of your data and your roadmap.
Agenda
18:00 — Doors + drinks + food
18:30 — Welcome
18:35 — Talks + Q&A
20:00 — Drinks + networking
21:00 — Close
About the speakers
Tobie Morgan Hitchcock — Co-Founder & CEO, SurrealDB
Tobie is CEO & Co-Founder of SurrealDB. An experienced tech entrepreneur and software engineer with 17 years in software and cloud computing, he specialises in distributed databases and highly-available systems. He founded SurrealDB in 2021 with the aim of building the ultimate cloud database for tomorrow's applications.
Martin Schaer — Solutions Engineer, SurrealDB
Martin is a computer science engineer working at SurrealDB and his GenAI startup. He recently worked in lab automation, designing declarative frameworks for instrument drivers and 3D visualisers for robotic transport. His background spans engineering, web development, and founding a successful agency in Costa Rica.
Ziv Ilan — Super AI team, NVIDIA
Ziv Ilan is a solution architect at NVIDIA, focusing on LLM evaluation, optimization, and deployment. He joined from deci.ai (acquired by NVIDIA), where he worked on neural architecture search and model compression techniques for generative AI and computer vision models. Ziv holds an electrical and electronics engineering B.Sc. from Tel-Aviv University and an M.B.A. from HEC Paris. He uses his technical and business expertise to drive AI solutions development and deployment.
About Agents in Production
Agents in Production is SurrealDB's regular meetup series for engineers shipping agents in 2026. Each edition takes one problem that's causing real pain in production and works through how teams are actually solving it.