Apache Superset

Preset is a Bronze sponsor of dbt Summit 2026. Stop by booth 337 and meet our team: - Maxime Beauchemin, Founder & CEO - Zane Aitken, Head of Sales - Vitor Avila, Sales Engineer - Evan Rusackas, Head of Community - Anna Lee, Marketing & Community Ops We’ll be talking shop on all things Apache Superset™ + dbt Labs while running live demos of Preset. See you there! Learn more and register: https://hubs.li/Q04xzCF60 #dbtSummit #preset #aianalytics

If you're going to #dbtSummit next week, put booth 337 on your list. The Preset team will be in Las Vegas September 15-18 running live demos of Apache Superset™ on dbt Labs. It's an AI-native, open analytics platform that reads the models and metrics you already maintain in dbt, so your dashboards stay in sync with your transformations without a second semantic layer. Stop by if you want to see a dbt project transformed into dashboards and insights, talk through what the future of AI analytics, or meet the people who build and run Preset. There might be some fun prizes in store, too. September 15-18 in Las Vegas: https://hubs.li/Q04xb_9y0 #dbtSummit #preset #aianalytics

Preset is proud to be a Bronze sponsor of dbt Summit 2026! We'll be in Las Vegas next week for the largest gathering of dbt Labs users anywhere: thousands of data practitioners and leaders, 100+ sessions, and three days of hands-on learning. Come find the Preset team at booth 337. We'll be showing how teams build AI-native, open analytics powered by Apache Superset™ directly on their dbt models, so the metrics defined in dbt are the same metrics people see in their dashboards, inside Preset Chatbot, or through our MCP. Details and how to find us at https://hubs.li/Q04wYRcx0 #dbtSummit #preset #aianalytics

Thank you, Airflow Summit 2026! To the venue teams, conference organizers at Software Guru, and everyone who stopped by to see Agor at our booth, we are grateful! Getting to host Agor in Austin was a blast, and the people who stopped to talk shop made the week. Maxime Beauchemin, Founder and CEO at Preset and Agor, is the original creator of Apache Airflow® and Apache Superset™. Airflow began as a way to orchestrate complex data workflows. Agor takes that same idea into agent orchestration: it’s a multiplayer canvas where you run agents the way you once orchestrated pipelines. Congratulations to our prize winners: 🎸 Grand prize, a Fender Stratocaster: Aldo Escobar, Data Engineer at Roxom 🎶 Second prize, a Moog Theremin: Atam Prakash Agrawal, Principal Engineer at BMC Software 🎧 Third prize, a Busy Bar: Charles Yu, Software Engineer at Datadog Thank you for making our booth a highlight of the week! See what we are building at https://lnkd.in/eBJAANaN

"Ask AI about your data" only works if your tools agree on what the data means. Ask three tools what "revenue" means and you might get three different answers. Point an AI agent at the same question and it reports whichever one it lands on, full confidence, no paper trail. The solution is upstream, in how the metric gets defined. Define it once in Preset’s semantic layer and every consumer reads the same definition: your dashboards, your SQL, and whatever LLM you connect. SIP-182 gives semantic layers a first-class connection type in Apache Superset, alongside databases, shipping in 7.0. dbt MetricFlow and Snowflake Semantic Views work on Preset today. Replay our deep dive: https://lnkd.in/eM9rrTEb Blog: https://lnkd.in/eip-vcX4

Our monthly community town hall is this Friday, August 28 at 9am pacific. It's an open call for anyone who uses or is curious about Apache Superset. We'll share what the community has been working on, walk through recent updates, and leave plenty of room for your questions. Whether you've shipped dashboards for years or just spun up your first chart, there's a seat for you. Join us Friday and find the calendar link here: https://hubs.li/Q04vf8Lt0

There will be Superset talks... you can count on it ;)

July was a busy month across the repo. 664 PRs, 132 contributors, 47 of them brand new. A few highlights our community shipped: • A unified Subject access model, so Users, Roles, and Groups work the same way for permissions • Dashboard-level async Excel export, plus PNG/PDF/JPEG export for charts • Soft delete with restore for charts, dashboards, and datasets (behind a flag) • Backend theme tokens to customize the UI from config, no fork required • OAuth 2.0 for Databricks and a new Apache DataFusion engine spec over Arrow Flight SQL • A large AI-assisted language translation effort, with dozens of locales backfilled using the repo's own tooling 74,200+ GitHub stars and 78 new folks in the community Slack. Thank you to everyone who contributed code, reviews, translations, and bug reports. Check out the full recap by PMC member Evan Rusackas. Link in comments below!

SIP-182 just landed on Superset master. Written by PMC member Beto Dealmeida, it gives semantic layers a proper connection type: peers of databases rather than pseudo-databases disguised as Postgres. Evan Rusackas's write-up covers the full design: how the Explorable protocol works, what SemanticView adds alongside Dataset and Query, and where Apache Ossie fits in as the vendor-neutral interchange format now under ASF governance. The semantic layer interface is deliberately small. The stated goal is that writing a provider for a new semantic layer should be a weekend project, not a quarter-long integration. If you want to contribute: Explorable is stable in master, the feature flag is SEMANTIC_LAYERS (targeted at 7.0), and the SIP itself is the right starting point: https://hubs.li/Q04sbTM-0 More at: https://hubs.li/Q04sbZf50 #apachesuperset #opensource #semanticlayers

The last thing you want in any analytics tool is trust issues. Ask three tools what "revenue" means and you'll get three answers, and AI will happily pick one at random. Defining metrics once in a semantic layer makes it so that every data consumer (dashboards, SQL, or [insert your favorite LLM here]) works from the same trusted definitions. Replay our latest live session featuring Evan Rusackas, special guest Beto Dealmeida, and Elizabeth Thompson for more: https://hubs.li/Q04rS5Cc0 #semanticlayer #apachesuperset #aianalytics