Schedule

In this talk, the speaker will look back at how Postgres grew from a small project into a top choice for building apps today. He will then look ahead at what Postgres 19 brings, why it is becoming popular for AI work, and what challenges lie ahead.

In this talk, the speakers will explain why PostgreSQL, not a mix of Redis, JSON files, and separate vector stores, should hold an AI agent’s memory and state. They will show how JSONB, pgvector, row level security, and transactions can work together in one schema, and discuss where a database still is not the right fit.

In this talk, the speakers will explain how PL/pgSQL actually runs behind the scenes, which helps explain why code that seems fine can turn slow or buggy later. They will cover loops, settings that affect performance, caching issues, common mistakes with exceptions, and other traps, shown through real code examples.

In this talk, the speaker will cover a Postgres 17 feature that keeps logical replication slots in sync on standby servers, so subscribers can keep going after a failover without needing a full resync. He will also show the setup steps and how to check that it works.

In this talk, the speaker will share a first hand account of moving a national tax system to PostgreSQL. He will cover checking against government rules, studying the old system, planning the new setup for reliability, and balancing performance and security with AI features still being worked on.

In this talk, the speaker will show how a certain type of lock can silently limit performance under heavy write traffic, based on real benchmark tests. He will explain how to spot this problem and how table partitioning can fix it, along with tips on choosing the right amount of partitioning.

In this talk, the speakers will describe how a large satellite and telecommunications company moved from Oracle to PostgreSQL across 16 countries while keeping Oracle as a safe backup using two-way data replication. They will cover how conflicts were avoided, how failures were handled, and how both databases were kept in sync.

In this talk, the speakers will look at what really happens when a Postgres extension is upgraded, and why upgrade scripts often fail in production. They will cover common mistakes, how Postgres picks an upgrade path internally, and share a checklist for building extensions that upgrade safely.