Aiven Blog

Mar 28, 2025

Aiven at pgDay Paris 2025

My first dive into the PostgreSQL® community was a great experience at PGDay Paris. Here is a recap, highlighting key talks on community, full-text search, lightning talks, Kafka & Debezium integrations, and more.

Tibs (Tony Ibbs)

|RSS Feed

Developer Educator at Aiven

pgDay Paris is a one day PostgreSQL® conference in Paris. This was the 9th time they've run pgDay Paris, and it showed in the quality of the conference. Aiven was proud to sponsor the event.

This was my first PG Day (indeed, my first PostgreSQL conference). It's not a large event, but it has a very strong sense of community, and the talks were of a very high quality.

And of course I got my photo taken with Slonik (I was not expecting that!)

Talks that stood out

The schedule is here.
Slides should be up for at least some of the talks. I think the talks were recorded, but am not sure.

Tuning community parameters: Hosting PostgreSQL events that perform

Ellyne Phneah of DBtune

Slides for the talk

The talk started with an explanation of the power of the open source community for PG - in particular the way that it allows a large group of people to work together, taking advantage of that expansion in knowledge and views.

It may not be an exact quote, but I wrote:

  • " PG is a community, not a technology, and this is its strength"

The talk was then the story of how the Malmo, Sweden PG user group was started (they had 45 people at their first meeting, which is very good), followed by advice on how to setup, run and publicize meetups. As someone who used to run a meetup, I can vouch for the excellence of the advice - there were several things I wish I'd been told back then!

If you have someone relevant visiting Malmo or Copenhagen (just across
the bridge!) then consider reaching out to the Malmo PG User Group (M-PUG) and offering a talk.

Full-Text-Search Explained from A to Z with French Food

Matt Cornillon, Google

I was quite excited about going to this talk, as text processing and search is something I find fascinating. As it turned out, if I had to choose one talk as my favourite at the conference, this would be it.

It was a very well structured and presented overview of "traditional" text search in PG, then of semantic search, and finally of how one combines the two. I took lots of notes!

Broadly, the talk covered:

  • Use of LIKE with wildcards (LIKE %croque-monsieur%) is inefficient
  • Full text search, which has been around in PG since 8.3.
    • How to make a ts_vector, with common root, stemming, suffix removal, and so on (which I knew how to do in OpenSearch but not in PostgreSQL).
    • Useful functions like to_tsvector and plainto_tsquery.
    • Various operators, including FOLLOWED_BY, and the @@ SQL operator.
  • Ranking with ts_rank, to give a better ordering to search results.
  • Mention of indexes, dictionaries of synonyms, pg_trgm and fuzzystrmatch (which I knew about), and weighting of results.
  • Then an introduction to semantic search, using the example of "seasonal fruits" as a phrase that has meaning beyond its parts. This I did know about from work at Aiven, for instance Building Smarter Documentation Search with Aiven for AlloyDB Omni and Gemini
  • Lastly, how to enrich a search by using hybrid search with rank() combining the results with COALESCE.

Lightning talks

I know lighting talks as short talks limited to 5 minutes at most, and this
was the format used here. I also love lightning talks, as it's both a great
way of getting into speaking, and also typically gives an excellent showcase to people's enthusiasms. I wasn't disappointed.

Topics covered (I didn't generally get speaker names)

  • State of the PG RPMs ("oops") - a wonderful "70+ slides reduced to 5 minutes" version of a longer talk.
  • Challenge Transparent Data Encryption - I'm not sure I'd thought about TDE, but apparently Oracle and SQLServer think this is the way to go, and PG doesn't do it (although someone pointed out the pg_tde extension). TDE as a concept raised my hackles, and this talk supported that instinctual reaction.
  • A series of unfortunate experiences with backup - PG backup, not "normal" backup.
  • Using PG to find my next Karaoke song - Floor Drees and Boriss Mejías advertising the karaoke session for the evening, but also a fun story of trying to find songs that would work well at karaoke.
  • pgcap2latex - a useful Wireshark plugin to output PG wire protocol in LaTeX or (I think) markdown. Tool available soon.
  • How to play wordle in PostgreSQL - using PG to do wordle on the 65 applicable PG keywords. Fun and with audience interaction.

The lightning talks were inspiring and fun!

As an Aiven person I take pride in knowing we encrypt data at rest and in transit - it's sensible for a managed service to handle this, but TDE is not enough. Also, now I want to learn how to find a karaoke venue to pair with the discovered songs (perhaps using postGIS, which I'm looking for an excuse to use).

Why PostgreSQL people should really care about Kafka and Debezium?

Our own Dirk Krautschick gave an excellent technical explanation of how to use the Apache Kafka CLI to create Kafka on your local machine, and get CDC (change data capture) going from PostgreSQL into Kafka using Debezium.

All of the resources (including slides) are available on GitHub at https://github.com/dkrautschick/pg_kafka_cdc

I hadn't quite realised Dirk's position in the PG community, until I discovered that he was responsible for taking the Slonic costume home with him after the conference! More seriously,
he runs the PG User Group NRW (North
Rhine-Westphalia)
. As I said for M-PUG above, bear that meetup in mind if you're in the area.

I do like the Apache Kafka CLI for working with non-managed services, but of course for my daily work I reach for the Aiven CLI and Aiven’s ease of use in creating services and managing their connections.

Explore the role of PostgreSQL in GitLab ecosystem

Hana Litavská, Faculty of Information Technology, Czech Technical University in Prague

I really didn't know what to expect from this talk, and I was delighted to
find a combination of a detective story and code archaeology into how GitLab uses PostgreSQL, from outside using only open source data and repositories.

I think I understood at most 2/3 of what was discussed, but the scale of the
investigation was amazing.

Afterwards

After the main conference day, there were two social events.

The first was the official get together at a very nice Belgian beer place (good food too). I got to share time with amazing people including some Aiven Crabs of past and present.

And of course our traditional crab gestures.

And afterwards there was karaoke, which was great fun. Sadly, I didn’t get to use my learnings from the earlier lightning talk.

From looking at the schedules of previous PG Day events, I knew to expect high quality talks, and I definitely got that. I also found a friendly community, and I hope to get to more such events in the future. I’m intending to be at PGDay UK in September and I know Aiven is sponsoring and will be present at PGDay Chicago. Be sure to follow our Events page to keep up with where we are in the postgreSQL community.


Stay updated with Aiven

Subscribe for the latest news and insights on open source, Aiven offerings, and more.

Related resources