Somewhere in the multiverse, there is a version of you who attends every interesting talk at PGConf.EU.
Unfortunately, in this universe, PostgreSQL has parallel query execution. Humans don’t.
Last year, I created an Underground Map for Database Nerds to navigate PGConf.EU without getting lost in the schedule.
This year, the conference returns to Valencia, October 20–23, with five parallel tracks and the same old problem: too many good talks occupying the same space-time coordinates.
So, I’ve borrowed a little inspiration from the Marvel Multiverse.
Imagine three alternate versions of yourself attending the same PostgreSQL conference, each with completely different priorities.
🌍 Universe Earth-5432: The PostgreSQL User
Superpower: EXPLAIN ANALYZE
DBAs, DBREs, developers, architects and anyone whose applications depend on PostgreSQL.
You’re interested in what PostgreSQL can do, how it works, how to make it faster, and why it occasionally ruins your weekend.
📣 Universe Earth-1996: The PostgreSQL Advocate
Superpower: Converting Oracle users without any violence.
Community organisers, speakers, educators, contributors, evangelists and anyone spreading the PostgreSQL gospel.
You care about adoption, community growth, open source, innovation and making PostgreSQL everyone’s favourite elephant.
💼 Universe Earth-2013: The Decision Maker
Superpower: GRANT BUDGET
CTOs, CIOs, engineering directors, team leads and architects who decide what the company runs, buys, builds or migrates.
You care about risk, reliability, architecture, operational costs and answering the eternal question: “Why not just use the cloud?”
You may inhabit more than one universe. That’s perfectly legal. Unlike PostgreSQL superuser privileges, nobody is checking.
Now, let’s find out where your variant should be.
🗓️ Tuesday, October 20 — Avengers Assemble (And Immediately Disagree)
The first day presents some unusually difficult choices.
11:15–12:05 — With Great PostgreSQL Power Comes Great Responsibility
🌍 USER | SM1-B — You Shall Not YOLO: Model Your Damn Data
Lætitia Avrot reminds us that good database design starts long before the first query hits production.
A must-see for developers and DBAs alike. Because PostgreSQL can forgive many sins, but a terrible data model will haunt you through every migration, index and performance-tuning session.
📣 ADVOCATE | Audit 1 — Postgres for Everything, Until It Isn’t
Gülçin Yıldırım Jelinek and Petr Jelinek explore where PostgreSQL shines and where its limitations begin.
Every good advocate should know both. After all, promoting PostgreSQL everywhere is much easier when you know where not to promote it.
Because even the biggest PostgreSQL fan shouldn’t try to run a toaster on it.
💼 DECISION MAKER | Audit 1 — Postgres for Everything, Until It Isn’t
Gülçin Yıldırım Jelinek and Petr Jelinek challenge the idea of using PostgreSQL for absolutely everything.
An important architectural discussion for leaders weighing technology consolidation against specialised systems.
13:05–13:35 — The Raspberry Pi Incident
🌍 USER | Audit 1 — Is PostgreSQL On My Raspberry Pi Faster Than Your Cloud Instance
Chris Ellis asks a question that could make several cloud invoices deeply uncomfortable.
A must-see for performance enthusiasts.
📣 ADVOCATE | Audit 2 — WAIT FOR the Feature for Almost 10 Years
Alexander Korotkov tells a story about the long journey of a PostgreSQL feature.
A useful reminder that open-source development is more marathon than sprint. Sometimes a very long marathon.
💼 DECISION MAKER | Audit 1 — Is PostgreSQL On My Raspberry Pi Faster Than Your Cloud Instance
Because nothing terrifies a CTO quite like discovering that their expensive infrastructure might be losing to a computer smaller than their coffee mug.
Verdict: User and Decision Maker timelines converge on Chris Ellis. Highly recommended.
13:45–14:35 — The On-Call Civil War
🌍 USER | Audit 1 — Postgres On-Call Confessions: Worst Practices 101
A panel dedicated to PostgreSQL operational mistakes.
Come for the horror stories. Stay for the comforting realization that other people have also done questionable things to production.
📣 ADVOCATE | Postgres Tuning in the Age of AI: Context Is Everything
Ryan Booz from pganalyze explores what changes, and what stubbornly stays the same, when AI joins the PostgreSQL performance-tuning party.
For advocates, a chance to understand where PostgreSQL stands in the AI era, what its ecosystem is doing about it, and whether decades of database wisdom are still relevant.
💼 DECISION MAKER | Audit 1 — Postgres On-Call Confessions: Worst Practices 101
Think of it as free incident-management training, funded by somebody else’s production outages.
Take notes. Some of these lessons are probably cheaper than your next postmortem.
⚡ 14:45–15:35 — THE MULTIVERSE FRACTURE
Two fantastic sessions. Same time. Different rooms.
Even Doctor Strange couldn’t find a timeline where you attend both.
🌍 USER | Audit 2 — The Monday We Learned Who Maintains Our Backups
David Steele and Jan Wieremjewicz.
A session involving the creator of pgBackRest and a story about backup maintenance? That’s enough to earn a front-row seat.
If you are a DBA/DBRE then you might even sacrifice a brilliant security talk for this one.
📣 ADVOCATE | Audit 1 — Nobody Gets to Be Superuser (Until I Broke In)
Mehmet Ince takes us into PostgreSQL security.
Nothing attracts attention like a real security investigation. And nothing builds trust in open source quite like examining its security boundaries seriously.
💼 DECISION MAKER | Audit 1 — Nobody Gets to Be Superuser (Until I Broke In)
Database privilege boundaries are a serious risk-management topic.
The title alone should be sufficient motivation to reconsider the permissions of that application account someone created in 2017.
⚡ Timeline warning: The backup session and Mehmet’s security session are both must-sees. The User’s choice is particularly painful.
16:05–16:55 — Backup, Sovereignty and TimeTravel
🌍 USER | Audit 2 — pgBackRest in HA setups: deployment patterns that work
Stefan Fercot dives into pgBackRest deployment patterns.
Because the primary going down should not also mean your backup strategy has gone on holiday.
📣 ADVOCATE | Audit 1 — Time Traveling PostgreSQL
Stefanie Janine Stölting takes PostgreSQL on a trip through time.
An intriguing topic to showcase PostgreSQL’s versatility beyond ordinary CRUD. After all, convincing people that PostgreSQL can do more than store rows is half the fun of advocacy.
💼 DECISION MAKER | SM1-A — Beyond Data Residency: Building Sovereign Control from Infrastructure to Data
Stephan Festl and Tim Steward discuss infrastructure and data sovereignty.
Particularly relevant for European organisations discovering that data residency and actual operational control aren’t necessarily the same thing.
17:05–17:55 — The Final Tuesday Timeline
🌍 USER | SM1-A — workspace_id % N: War stories on building resilience through application-level sharding
José Neves discusses the adventures of application-level sharding.
Modulo arithmetic is easy. Distributed data ownership, resharding and operational resilience are somewhat less easy.
📣 ADVOCATE | Audit 2 — Community Moments That Shaped 30 Years of PostgreSQL
Valeria Kaplan takes us through PostgreSQL community history.
Because PostgreSQL’s greatest feature might be the people who kept developing it for three decades without a single corporate owner.
💼 DECISION MAKER | Audit 1 — A PostgreSQL 19 whirlwind tour
Jonathan S. Katz takes us on a whirlwind tour of what’s coming in PostgreSQL 19.
A perfect opportunity to discover new capabilities, spot opportunities for your technology roadmap, and see where PostgreSQL is heading next.
After all, the best way to predict your database’s future is to attend the talk before approving next year’s budget.
18:00–19:30 — PostgreSQL Europe Reception
All universes collapse into the hallway track.
🗓️ Wednesday, October 21 — The Multiverse of Madness
The honeymoon is over. Today we’re dealing with vacuum, monitoring, memory, migration and why SQL NULL behaves like a philosophical argument.
09:25–10:15 — Three Different Mornings
🌍 USER | Audit 2 — Autovacuum Is Running. Is It Actually Working?
Teresa Lopes tackles the eternal mystery of PostgreSQL maintenance.
Autovacuum running is good. Autovacuum achieving something useful is better.
📣 ADVOCATE | Audit 1 — Creating a “Dungeon Master” with Postgres and MCP
Matt Cornillon brings PostgreSQL into an AI-assisted application.
A potentially entertaining way to demonstrate PostgreSQL’s versatility to audiences who don’t dream about work_mem management.
💼 DECISION MAKER | Audit 1 — Creating a “Dungeon Master” with Postgres and MCP
A fun demonstration of how PostgreSQL can participate in the AI ecosystem without reinventing the entire technology stack.
Because sometimes the best AI strategy isn’t buying another database. It’s discovering what your existing one can already do.
10:45–11:35 — Documentation, NULL and Downtime
🌍 USER | Audit 2 — What is NULL?
Vik Fearing explains NULL.
Thirty years of SQL experience will not prevent somebody from getting three-valued logic wrong tomorrow.
📣 ADVOCATE | Audit 1 — How to Contribute to the PostgreSQL Documentation
David E. Wheeler shows how to contribute to PostgreSQL documentation.
Not all heroes write C. Some fix the documentation that prevents thousands of people from writing terrible SQL.
💼 DECISION MAKER | Audit 3 — When zero-downtime saves lives: the reality of Active-Active Postgres
Adam Wright and Boriss Mejias from EDB share the story of an emergency-response system running across island datacenters, where network partitions are common and downtime can genuinely put lives at risk.
A fascinating real-world lesson in Active-Active PostgreSQL, split-brain scenarios, conflict resolution and near-zero-downtime upgrades.
Because sometimes five nines aren’t about impressing your CTO. They’re about ensuring someone can reach emergency services.
11:45–12:35 — Diagnose, Improve, Migrate
🌍 USER | Audit 1 — Stop Guessing: Unlocking Wait Events at Scale with Existing and New Tools
Dmitry Fomin investigates wait events.
Because the technical explanation for “the database is slow” should ideally involve more than vibes.
📣 ADVOCATE | pg_hardstorage: Modern backups for PostgreSQL
Hans-Jürgen Schönig from CYBERTEC introduces pg_hardstorage, a new open-source contribution to the PostgreSQL backup ecosystem, designed for large-scale and regulated environments.
For advocates, another example of PostgreSQL’s ecosystem continuing to innovate beyond the core database.
💼 DECISION MAKER | Audit 2 — Oracle to PostgreSQL beyond the Syntax: When DBMS Design Differences Matter
Joshua Steinmann and Tino Engelbrecht examine migration differences.
Turns out replacing Oracle with PostgreSQL involves a little more than changing the connection string and cancelling the licence.
13:35–14:25 — Your Next Big Decision
🌍 USER | Audit 1 — The anatomy of PostgreSQL row locks
Laurenz Albe dissects PostgreSQL row-level locking, one of those things every developer and DBA thinks they understand until a transaction starts blocking half the application.
📣 ADVOCATE | SM1-A — Selling PostgreSQL?
Jan Karremans discusses PostgreSQL adoption and positioning.
A natural choice for anyone trying to spread the good word without turning into a database salesperson.
💼 DECISION MAKER | Audit 2 — Security Strategies for your database backups
Priyanka Chatterjee discusses backup security.
A backup isn’t a safety net if the same incident can compromise it too.
14:35–15:25 — Memory, Migration and Other Expenses
🌍 USER | Audit 1 — Zooming in on OOM: A Deep Dive into Postgres and Linux Memory
Dimitris Sideris and Oleksii Kliukin explore PostgreSQL memory behaviour.
Because Linux OOM killer has absolutely no respect for your uptime SLA.
📣 ADVOCATE | Audit 2 — Relational or graphs? Why not both?
Ants Aasma explores graph capabilities alongside the relational model.
Useful ammunition for explaining why every new data-access pattern does not necessarily require a new database.
💼 DECISION MAKER | SM1-A — Leaving Oracle and SQL Server for PostgreSQL: What Could Go Wrong
Avi Vallarapu and Michal Nosek examine database migration risks.
Licence savings are exciting. Migration surprises tend to be less so.
15:35–16:25 — It Worked on My Database. Famous Last Words.
🌍 USER | SM1-A — It Works on My Database
Radim Marek from boringSQL tackles the classic mystery: a query runs in 4 milliseconds in development but takes 12 seconds in production.
Learn how PostgreSQL lets you bring production planner statistics into your development database, without copying millions of rows or sensitive customer data.
📣 ADVOCATE | SM1-B — Same Problems, Different Defaults: What 25+ Years of SQL Server Taught Me When I Switched to PostgreSQL
After 25 years with SQL Server, Silvano Coriani from Microsoft explores what changes, what doesn’t, and which familiar assumptions stop working when switching to PostgreSQL.
A fascinating perspective for advocates trying to win hearts and minds beyond the PostgreSQL bubble.
Because converting someone from SQL Server is much easier when you understand their world, rather than just telling them PostgreSQL is better.
💼 DECISION MAKER | Audit 1 — What does the EU Cyber Resilience Act mean for PostgreSQL and its community?
Peter Eisentraut explores what the EU Cyber Resilience Act means for PostgreSQL and its ecosystem.
An important session for anyone responsible for open-source adoption, compliance and technology governance.
Because the only thing more complicated than database licensing is European regulation.
16:55–18:30 — Sponsor Keynote Infinity Stones
The regular tracks give way to five consecutive platinum sponsor keynotes in Audit 1.
Every universe gets the same options: PostgreSQL tuning, AI systems, Google Cloud, EDB’s history and Microsoft’s PostgreSQL strategy.
One can collect all five Infinity Stones without changing rooms. Whether this grants you control over production incidents remains untested.
🗓️ Thursday, October 22 — PostgreSQL: Endgame
The final conference day. By now your laptop has 37 browser tabs, your notebook has SQL fragments with no context, and your conference badge has survived more conversations than your connection pool.
09:25–10:15 — The Future Arrives
🌍 USER | Audit 1 — A demo only tour through PostgreSQL 19
Daniel Westermann demonstrates PostgreSQL 19 features.
The perfect session for people who prefer watching features work over reading 90 slides about why they’re exciting.
📣 ADVOCATE | Audit 2 — Finally a Postgres Communication Hub for Experts and Newcomers
Jack Bonatakis and Kai Wagner explore connecting members of the PostgreSQL community.
Because a thriving ecosystem needs more than GitHub issues and occasional arguments on mailing lists.
💼 DECISION MAKER | Audit 1 — A demo only tour through PostgreSQL 19
See what is coming and start thinking about the next upgrade cycle.
The database roadmap waits for no quarterly planning meeting.
10:25–10:55 — The Unsung Heroes and the Ancient Origins
🌍 USER | Audit 1 — radar + radar-ui: Agentless Diagnostic Snapshots for PostgreSQL
Jimmy Angelakos introduces diagnostic snapshots without the usual agent deployment overhead.
Because when production breaks, “we should have enabled more monitoring” is the least useful thing anyone can say.
📣 ADVOCATE | The process of RPM packaging in the PostgreSQL ecosystem
Devrim Gündüz takes us behind the scenes of PostgreSQL RPM packaging.
Millions of people install PostgreSQL with a simple yum install, but somebody has to make that magic work across distributions and versions.
Because not all PostgreSQL heroes wear capes. Some maintain RPM repositories so the rest of us can pretend installation is easy.
💼 DECISION MAKER | Audit 2 — Postgres Before PostgreSQL: Stories from Inside the Original Berkeley Lab
Ellyne Phneah takes us back to Berkeley in the 1980s, revealing the surprisingly human stories behind PostgreSQL’s earliest design decisions.
A fascinating lesson in the Lindy effect: technology that has survived decades of change has demonstrated remarkable staying power. PostgreSQL has outlived countless database fashions and still keeps evolving.
11:25–12:15 — The Database Has Entered Its Automation Era
🌍 USER | Audit 2 — How to detect and fix corruptions
Derk van Veen tackles corruption detection and repair.
Hopefully the only time you ever need these techniques is during a conference talk.
📣 ADVOCATE | SM1-A — Autonomous PostgreSQL optimization: End-to-end tuning for parameters, indexes, and autovacuum
Luigi Nardi explores autonomous database tuning.
An interesting look at how PostgreSQL management could become more accessible to people who don’t have fifteen years of DBA experience.
💼 DECISION MAKER | SM1-A — Autonomous PostgreSQL optimization: End-to-end tuning for parameters, indexes, and autovacuum
Can your database fleet grow without hiring one more DBA for every handful of instances?
Automation might be part of the answer. Understanding its limitations is equally important.
13:15–13:45 — PostgreSQL Logs, TikTok and a Little Comic Relief
🌍 USER | PostgreSQL Logging in Practice
Alicja Kucharczyk explores how to turn PostgreSQL logs from an endless stream of cryptic messages into something actually useful for troubleshooting.
📣 ADVOCATE | A composition of short items (or: a talk for the tiktok generation)
Ellert van Koperen breaks away from the traditional conference format with a collection of short PostgreSQL topics.
After three days of serious database discussions, here’s a chance to loosen up, enjoy something different, and give your brain a well-deserved context switch.
💼 DECISION MAKER | A composition of short items (or: a talk for the tiktok generation)
No grand architectural decisions, migration strategies or budget justifications. Just an opportunity to enjoy something different before returning to the corporate universe.
For once, you can enjoy a PostgreSQL talk without calculating its ROI.
13:55–14:45 — Performance, Planning and Architecture
🌍 USER | Audit 2 — AIO performance — now and in the future
Tomas Vondra discusses asynchronous I/O performance.
Storage performance remains one of PostgreSQL’s most interesting battlefields.
📣 ADVOCATE | Audit 1 — pg_plan_advice: Plan Stability and User Planner Control for PostgreSQL
Robert Haas explores planner control.
Another example of PostgreSQL evolving in response to real users who occasionally disagree with the optimizer.
💼 DECISION MAKER | Audit 1 — pg_plan_advice: Plan Stability and User Planner Control for PostgreSQL
Robert Haas explores pg_plan_advice, a potential PostgreSQL 19 feature that gives users more control over query execution plans.
A particularly exciting development for organisations migrating from Oracle, where execution plan stability and optimizer hints have long been part of the DBA toolbox. One less reason for Oracle veterans to say, “But PostgreSQL can’t do that!”
14:55–15:45 — The Final Boss
🌍 USER | Audit 1 — Identifying bottlenecks in Postgres workloads
Andres Freund discusses PostgreSQL bottleneck identification.
A fitting final boss for every performance enthusiast.
📣 ADVOCATE | Audit 1 — Identifying bottlenecks in Postgres workloads
Yes, advocates need technical substance too.
Being able to explain PostgreSQL’s strengths honestly is far more valuable than waving another benchmark graph around.
💼 DECISION MAKER | Audit 3 — CPO: Multi Data Center Kubernetes on steroids
Hans-Jürgen Schönig explores multi-data-centre Kubernetes-based PostgreSQL operations.
A session for decision makers weighing operational complexity against flexibility and resilience.
15:55–16:25 — Three Final Portals
🌍 USER | Audit 3 — Finding the invisible: spotting query planner issues before they appear
Andrei Lepikhov explores identifying planner problems.
Because the best production incident is the one that never happens.
📣 ADVOCATE | Audit 2 — Acknowledging PostgreSQL Contributors: The Contributor Badges Project
Christoph Berg and Melanie Plageman explore recognising PostgreSQL contributors.
Open source runs on contributions. A little recognition probably doesn’t hurt either.
💼 DECISION MAKER | SM1-A — What Actually Goes Wrong with Postgres on Kubernetes
Natalia Marukovich explores PostgreSQL-on-Kubernetes failure modes.
An excellent reality check before signing off on a platform strategy consisting entirely of YAML and optimism.
16:55–17:40 — Lightning Talks
All timelines converge once again.
Short talks, unexpected ideas and possibly somebody trying to explain a new extension in five minutes.
🗓️ Friday, October 23 — The Post-Credits Scene
The official conference talks are over, but Community Events Day offers one final adventure.
🌍 USER: PostgreSQL Observability Summit in the morning, PostgreSQL on Kubernetes Summit in the afternoon.
📣 ADVOCATE: New PostgreSQL Community Members Summit or community education events in the morning, then PG Outreach Unconference or Storytelling for Technical Communication in the afternoon.
💼 DECISION MAKER: A Postgres Ecosystem Foundation? or Autonomous Postgres in the morning, then the Kubernetes Summit to discuss the reality of database platform operations.
Community Events Day requires separate registration. Not all events are open in the same way, so check their individual participation details.
🐘 The Endgame: There Is No Perfect Timeline
Three universes. Three very different ways to experience one conference.
The User wants a faster, safer, more reliable PostgreSQL.
The Advocate wants PostgreSQL to conquer the world.
The Decision Maker wants all of that, preferably by the end of the quarter and without increasing the budget.
Unfortunately, no Infinity Stone can fix scheduling conflicts.
But that’s okay.
Because the best part of PGConf.EU has never been just the scheduled talks. It’s the conversations afterwards, the arguments over coffee, the people who have already solved your weird production problem, and occasionally the people who helped create it.
See you in Valencia!
Official PGConf.EU 2026 Schedule
Disclaimer: These are personal, opinionated picks, nothing official about them. No parallel universes were harmed in the creation of this guide. Your production database, however, remains your responsibility.
🌀 Beyond the Multiverse: The PostgreSQL Hackers
And then there are the PostgreSQL hackers and contributors.
These mysterious interdimensional beings exist outside our three universes. They don’t need a conference survival guide. They can jump between query planners, WAL internals, community governance, and performance tuning without suffering a context switch.
Some of them even wrote the code everyone else is attending talks to understand.
While the rest of us explore the PostgreSQL multiverse, these are the people who keep adding new dimensions to it.
🎉 Friday, October 23 — The Multiverse Afterparty: Community Day
Three days of slides, SQL, benchmarks and architectural debates. Time to loosen up.
Community Day is where PostgreSQL gets more interactive, informal and collaborative. Think unconferences, open discussions, workshops, brainstorming and meeting the people who make the elephant tick.
No need to choose your universe today. Users, advocates, decision makers and even those interdimensional PostgreSQL hackers can mingle freely.
And remember: What happens at Community Day stays at Community Day. 😉 Especially since no recordings have been announced. If you want to know what happened, you’ll probably have to show up.
