A one-day database event like no other—live or virtual.

This workshop introduces the foundational concepts behind CockroachDB as a distributed SQL database designed for resilience, scale, and consistency. The focus is on understanding how CockroachDB differs from traditional databases—and why its architecture enables survivability, horizontal scaling, and strong consistency across regions without manual intervention.

Key Topics:

- Distributed SQL fundamentals: nodes, ranges, replication, and rebalancing
- The Raft consensus protocol: how CockroachDB ensures consistency and fault tolerance
- ACID transactions in a distributed system: serializable isolation by default
- Automatic data distribution and horizontal scaling without manual sharding
- Survivability and high availability: handling node, zone, and region failures
- Multi-region fundamentals: latency, locality, and data placement strategies
- SQL compatibility and PostgreSQL wire protocol basics

Leadenhall Hub 1

This workshop covers vector storage, indexing, and retrieval-augmented generation (RAG) using CockroachDB as a unified operational and vector database. The focus is on why ACID transactions and referential integrity change the game for AI applications—eliminating orphaned vectors, ghost data, and consistency gaps that plague dual-database architectures (PostgreSQL + Pinecone).

Key Topics

- Why ACID matters for AI: transactional ingestion, atomic updates, cascading deletes—no orphaned vectors, ever
- CockroachDB VECTOR type and C-SPANN indexing internals
- RAG pipeline design with transactional guarantees: all-or-nothing ingestion
- Hybrid search: combining vector similarity with SQL predicate pushdown and JOINs
- Multi-tenant vector isolation: using SQL predicates and foreign keys
- Performance benchmarking: C-SPANN vs. brute-force at scale

Leadenhall Hub 2

Matt Gardner sits down with Jordan Legg, Chief AI Officer at Takara AI, to talk vectors, scale, and what CockroachDB + DS1 make possible — then wraps it up with a rapid-fire This or That. Swing by the Crawl Space stage for your front-row seat.

Cockroach Labs Staff Sales Engineer Anna Eriksson talks to more engineering teams in a month than most people do in a year. In this session, she and Matt pull back the curtain on the patterns she sees in the field: the decisions that hold teams back and the ones that quietly save the day. Stick around for Red Flag / Green Flag where she judges architecture decisions in real time.

For eleven years, CockroachDB has made the impossible possible, surviving regional outages with full consistency, transacting across continents, scaling elastically without re-architecture. Possible, but never easy. This year, that changes.

Each year at RoachFest, Spencer previews the work ahead. This year's preview is the most consequential of CockroachDB's history: the new database estate with a unified architecture operated by a continuous AI team for every cluster. He will show why the database that already runs your most critical workloads is becoming the database that runs every workload at a cost basis no other database can match and why the agents that build, fill, and operate the modern enterprise need a database they themselves can run on, and run.

The agentic database cloud begins this year, built on the database you already trust.

Spencer Kimball, Co-Founder & CEO
^{Main Stage | Bishopsgate Forum}

Most disaster recovery plans are tested in staging. Form3 tests theirs in production. In this session, Kevin Holditch, VP of Engineering at Form3, shares how the company evolved from a traditional AWS-based platform into a cloud-agnostic, active/active/active architecture spanning AWS, GCP, and Azure using CockroachDB, Kubernetes, NATS, and Go. The platform is designed to survive the loss of an entire cloud provider without customer impact. Kevin will walk through the engineering decisions, operational tradeoffs, and lessons learned building resilient payments infrastructure for tier-1 banks — including how Form3 validates disaster recovery by intentionally shutting down a cloud provider for 24 hours in production.

Kevin Holditch
^{VP of Engineering, Form3}

_{Main Stage}

Everyone agrees AI agents need memory. The harder question is what “memory” actually means in production systems. Should long-term state live in a vector store or a relational database? Should memory be shared across agents or isolated per user, workflow, or task? And how do you keep that state accurate as the world changes underneath it? This panel brings together experts from Memori Labs, Takara AI, and Cockroach Labs who build and operate agent systems at scale. Panelists will discuss the tradeoffs behind semantic retrieval, SQL-native memory layers, distributed state management, retrieval architectures, and always-on infrastructure for persistent agents. The conversation will also explore some of the most practical enterprise challenges in agent memory systems: reducing token and inference costs through durable context, avoiding vendor lock-in with model-agnostic memory architectures, and structuring memory from agent traces so systems can improve over time instead of simply accumulating context. Attendees will leave with a practical understanding of how to build durable memory systems for AI agents that remain consistent, scalable, observable, and useful across long-running and multi-agent workflows.

Adam B. Struck
^{CEO & Co-Founder, Memori Labs}

Jordan Legg
^{Chief AI Officer, Takara}

Rob Reid, Techical Evangelist
^{Cockroach Labs}
_{Main Stage}

Spill the tea with Spencer Kimball

Spencer Kimball built a database named after one of the world’s most resilient — and least appreciated — creatures. Now he’s spilling the tea.

In this no-slides, no-filter session, CockroachDB’s CEO sits down for hot takes, hard questions, and opinions that usually stay off the keynote stage. From AI hype cycles to database drama to the comments people still leave about the name, nothing is off the table. Well… except the tea.

Spencer Kimball, Co-Founder & CEO
^{The Crawl Space Stage | Bishopsgate Forum}

The teams building the world's most important applications need a database that evolves with them, one that's ready for where they're going, not just where they've been. We'll dig into new product innovations such as our Bring Your Own Cloud (BYOC) which gives enterprises the fully managed experience they rely on, running inside their own infrastructure. We'll showcase our AI-powered migration tooling, making the move to CockroachDB faster and lower-risk than ever. And we'll share our approach to making CockroachDB agent-ready, built for the way modern AI applications are being designed and deployed. Looking ahead, we'll also share our approach for making CockroachDB the most price-performant database at scale (i.e. disaggregated storage and compute, auto-scaling, and offloading workload management), and how we're thinking about the future demands of AI data infrastructure.

Igor Stanko, VP of Product
^{Main Stage | Bishopsgate Forum}

Tim Peake is a former Army Air Corps officer, test pilot, and the European Space Agency’s first British astronaut to travel to the International Space Station. Selected from more than 9,000 applicants for ESA’s astronaut program, Tim spent six months aboard the ISS as part of the Principia mission, where he conducted more than 250 scientific experiments, performed a spacewalk, and helped engage millions around the world through live broadcasts and educational outreach from orbit.

Since returning to Earth, Tim has become one of the world’s most recognizable advocates for science, technology, and innovation. Drawing from his experiences in space and the military, he speaks on leadership, teamwork, risk management, resilience, and decision-making in high-pressure environments. He is also the bestselling author of several books on space exploration, STEM, and the future of science and technology.

We’re excited to welcome Tim Peake to the RoachFest stage.

Major Tim Peake CMG
^{Main Stage | Bishopsgate Forum}