PGConf Brazil is one of the highlights of the Postgres calendar. This year from September 3 to 5, João Pessoa will host three days of conversation, problem-solving, and big ideas from across Latin America’s thriving Postgres ecosystem.
EDB is proud to be back as a Gold sponsor. EDB event organizers Euler Taveira, João Foltran, and William Ivanski have helped shape the program, and our Brazilian and global colleagues are bringing a brilliant lineup of talks and workshops ranging from database internals to AI, high availability, and performance tuning.
The 2025 Stack Overflow Developer Survey once again ranked Postgres the most used, most admired, and most desired database. That recognition belongs to the people building, running, and sharing it every day. We’re proud to be part of that work, and this year’s PGConf Brazil program is a perfect example of it.
Keynote
Bruce Momjian — The Postgres Trajectory
EDB Vice President, Postgres Evangelist, and Core contributor Bruce Momjian will look at how Postgres has evolved, where it stands today, and where it’s headed. Drawing on decades of experience and travels to Postgres events around the world, Bruce will share insights into the project’s growth, its role in the modern data landscape, and the forces shaping its future.
Core Database Internals and Performance
Euler Taveira — Desmistificando VACUUM (Demystifying VACUUM)
Postgres depends on regular maintenance to deliver consistent performance, and VACUUM is at the heart of that process. This session covers how VACUUM and VACUUM FULL work, how autovacuum operates beyond simple scheduling, and the risks of neglecting them — from query slowdowns to transaction ID wraparound. Euler will share strategies for diagnosing VACUUM-related issues, optimizing configurations, and monitoring key indicators like table bloat and transaction ID consumption to prevent production problems before they start.
Juliano Atanazio — JIT: aliado ou não? (JIT: Ally or Not?)
Just-In-Time (JIT) compilation can accelerate complex queries by compiling parts of the execution plan in real time, but it isn’t a silver bullet. This talk explains exactly how JIT is triggered, when it should be enabled or disabled, and how to configure it effectively. Expect practical benchmarks, workload-specific recommendations for OLTP and OLAP environments, and guidance to avoid performance regressions caused by indiscriminate use.
Barbara Leidens — Entendendo o Funcionamento do Query Executor (Understanding the Query Executor)
From a query’s logical plan to its physical execution, Barbara will walk through the complete lifecycle inside Postgres’ Query Executor. Topics include recursive traversal of execution trees, operator chaining in scans, joins, aggregates, sorts, and limits, and how resources are consumed along the way. She will also cover expression evaluation, the “pull” model of tuple streaming, and how the executor leverages intra-query parallelism, including optimizer decision criteria like cost, selectivity, and data size.
Gustavo Oliveira — PostgreSQL locks: tipos e usos (PostgreSQL Locks: Types and Uses)
Concurrency in Postgres is enforced through a rich set of locks. This session introduces the different lock types, when they are acquired, and how they can impact query performance. Learn to identify active locks in real time, interpret the lock-related information exposed by Postgres, and diagnose contention issues to maintain consistency without sacrificing throughput.
Andre Marchesini De Aguiar — Armazenamento no PostgreSQL (Storage in PostgreSQL)
Data storage affects everything from query performance to disk usage, and Postgres’ approach is more complex than it might seem. This talk examines how data is organized into files and 8KB pages, how tuples are laid out within blocks, and the role of auxiliary files like .fsm, .vm, and .init. Andre will also explain the TOAST mechanism for large attributes — how it fragments, compresses, and retrieves oversized data — and show how tools like pageinspect can reveal internal structures, diagnose bloat, and guide best practices for large-field management.
High Availability and Operations
Charly Batista — Alta Disponibilidade: Domine Failover, Load Balancing e Backups (High Availability Workshop)
This four-hour, hands-on workshop covers the full stack for high availability in Postgres: Patroni for cluster automation and failover, ETCD for distributed coordination, HAProxy for load balancing, and pgBackRest for robust backup and restore. Attendees will learn architectural fundamentals, advanced configuration strategies, disaster recovery planning, and testing techniques to validate performance and resilience.
Charly Batista — Reducing Costs and Improving Performance With Data Modeling in Postgres
By understanding how Postgres organizes data within blocks and leveraging the Free Space Map (FSM), it’s possible to reduce storage footprints, improve memory efficiency, and optimize performance. This talk explores strategies for reorganizing the data model to align data more efficiently, with practical examples of how these changes can deliver both cost savings and measurable speed gains.
Israel Barth — Understanding Patroni: A High Availability Solution for Postgres
High availability is a priority in any critical database environment, but implementing it does not have to be complex. This session introduces Patroni — one of the most accessible and effective HA tools for Postgres — and explains how it works, from the fundamentals of HA to the specifics of leader election, failover coordination, and maintaining cluster consistency. Israel will cover the role of each component (Patroni, Postgres, distributed configuration store, and proxy), walk through setting up a functional cluster, and share practical tips for operating it safely in production. Attendees will leave with a clear understanding of Patroni’s capabilities and a foundation for building resilient Postgres clusters without unnecessary complexity or vendor lock-in.
Euler Taveira — Security in PostgreSQL (Workshop)
A practical workshop on securing Postgres installations. Topics include authentication and authorization, roles and permissions, configuration best practices, and advanced security controls to protect sensitive data and maintain compliance in production environments.
Advanced Features and Emerging Workloads
Bilge Ince — From Distance to Intelligence: Evolving Your Vector Search
Full-text search and vector search each have strengths and limitations. This session explores how to combine them — and other Postgres capabilities — to deliver more accurate, context-aware search results. Bilge will cover the trade-offs of each approach, why vector distance alone can miss relevant matches, and how to build search systems that balance semantic relationships with exact-match precision for production workloads.
Community and Ecosystem
Israel Barth — De usuário a contribuidor (From User to Contributor)
Contributing to Postgres is rewarding, but the process can seem daunting to newcomers. This talk breaks down the full development workflow, from release cycles and Commitfest participation to patch submission, code reviews, and testing. Israel will walk through setting up a development environment, compiling Postgres, running tests, and following the project’s style and quality guidelines. Attendees will leave with practical steps for making their first meaningful contribution.
Raphael Vieira — WarehousePG: Greenplum de código aberto (WarehousePG: Open Source Greenplum)
WarehousePG is EDB’s Apache-licensed fork of Greenplum, designed to give organizations an open-source path for massively parallel processing (MPP) data warehousing. Raphael will cover the history of Greenplum, the architecture of WarehousePG, and key features like “Shared Nothing” segment nodes, master-standby configurations, table distribution methods, heap and append-optimized tables, columnar storage, compression, the GPorca optimizer, and backup utilities. The session also addresses how existing Greenplum deployments can transition smoothly to WarehousePG.
See You in João Pessoa
Whether you’re there to learn new skills, dig into technical deep dives, or connect with others who care about Postgres, PGConf Brazil is worth making time for. We look forward to the conversations, ideas, and collaborations that will come out of this year’s program.
Want to work on some of the same challenges we’ll be discussing in João Pessoa? Explore our open roles and see how you can help us take Postgres even further.