i have never posted the rocket ship emoji

A North American river otter has been documented along the Bronx River for the first time in approximately 100 years! (1/4) 📸: Wildlife Conservation Society Education Department / RJ Hawkins

8,000+ Dags across 100+ engineers...@Wix knows all about about orchestration at scale. At Orchestrate Everything, Yarden Wolf will walk through how they built a context-aware agentic coding platform on top of that scale, and how it's saving their team 850 engineering hours a month. One of five sessions of real production AI stories from companies like Ramp, Lyft, Mercor, and Aampe. Sept 16. Virtual. Free. Register now: astronomer.io/events/orchest… #dataengineering #AI #Airflow #virtualconference

A Black-crowned Night-Heron stands over the shore of the Central Park Lake.

A great trip depends on perfect timing. So does the AI that plans it. Using Astro, @bookingcom orchestrates the data, tools, and LLMs behind its Connected Trip vision and the AI suite delivering it, from the AI Trip Planner to the agentic Smart Messenger. Here's how it works: Every AI assistant a traveler talks to depends on data arriving at the right place, right on schedule. Astro runs that itinerary: pipelines prepare traveler data, build the context each assistant draws on, refresh agent memory, and score how the assistants respond in production. Getting there took one of the largest data platform migrations in European history. Thousands of pipelines moved off self-managed @ApacheAirflow in a matter of months, with zero disruption to the systems behind bookings, payments, and partner payouts. With Astro, Booking has: ✈️ Hundreds of GenAI data pipelines in production 🧑‍💻 Thousands of data practitioners building on Astro ⚡ Hundreds of thousands of task runs per day 🛡️ Near-zero scheduler downtime (previously days per year) Read the full case study: astronomer.io/case-studies/b…

Carnegie Mellon University's Delphi Group forecasts epidemics. When COVID pushed their systems past what they were built for, they made a deliberate call afterward to rebuild it as something open, modular, and designed to outlast any one crisis. The requirements were specific: aggregate signals across geography and time, revisit and revise data as it comes in, and keep it open source so the work can be checked, not just trusted. That's what led them to Airflow, and Astro to run it, porting years of pipelines built during the emergency onto a platform designed to handle exactly that kind of complexity. Serious infrastructure looks like this in epidemiology. It should look like this in AI too. Check out @diginomica's piece on the rebuild: bit.ly/4bjzY5l

Over the weekend a school of rays were spotted in the Newtown Creek #whatisnewyork

Last year, 8,000+ data engineers tuned into our first virtual conference, Beyond Analytics, to learn from real practitioners how they use data orchestration to power way more than dashboards. The scope of what orchestration drives has only continued to expand in the AI era, so on Sept. 19 we're running it back with a new name: Orchestrate Everything. This conference is all about data leaders from top brands that are orchestrating the foundation of AI that’s actually in production. Here’s what to expect: 🧠 Marc Lamberti’s Airflow AI Orchestration certification crash course + free cert code ($150 value) 🔧 Real production stories from real data teams: AI inference, MLOps, and context engineering (Hear from @tryramp, Aampe, and more! ⚡ Deep dive into how Astronomer is carrying its data orchestration foundation into AI Register now: astronomer.link/urwRF #dataengineering #ai #virtualconference

🤖 Data pipeline needs grow faster than headcount. @togethercompute's answer: AI agents build the pipelines on Astro. Pipeline development is now 20x faster. How it works: An engineer submits a ticket, an agent reads the ticket, authors the Airflow DAGs and dbt models, tests live on Astro via the Airflow MCP, and opens the PR. Humans review. Astro ships it. ⚡ 700+ agent-authored dbt models in production. Deploys in seconds, not minutes.

"The pressure that using Astronomer took off our cloud infrastructure team was immense. I consider Astronomer one of the non-negotiables." — Bill Dunaway, Senior Director of Data Engineering and Operations, @ClaycoInc 🏗️ Clayco is a $7.5B design-build firm with 50-100 active job sites at any given time. Their data engineering team knew exactly what they needed: reliable orchestration that didn't require them to become cloud infrastructure experts to maintain it. Clayco chose Astro. In under 90 days, the team consolidated 4 legacy schedulers into a single platform. ERP syncs went from an hour to 15 minutes. Dag count grew 20x in the first year. Case study link below.

✍️ Carter Page, EVP of R&D at Astronomer, writes in @CIOonline about why so many early AI initiatives are struggling, not because models are weak, but because the context they need to operate reliably inside organizations is missing. 💭 "Orchestration is typically the one layer that already has visibility across platforms, spanning your warehouse, your transformation layer, your external APIs, and your operational databases. That cross-platform vantage point is what makes it possible to build a context layer that is comprehensive rather than siloed." Read the article for Carter's advice on the 2️⃣ key investments that organizations deploying AI agents at scale need to make, before their agents reach production.

I was sitting on our Morningside Heights roofdeck with my camera when suddenly this American Kestrel flew by, practically at eye level. It made a quick turn and stared right at me before flying off. What a view of its gorgeous feathers in flight! #birdcpp

A rare American Bittern rests on Duck Island on Central Park's Harlem Meer today.

A two-task plugin routing Slack pipeline that alerts to predefined commands or an AI token. Samantha Blaney Cuevas of @JeppForeFlight shares on building the Data Port agent in Airflow. Links below. #AI #Automation #Airflow

👋

The Kwanzan cherry blossoms are now in full bloom in Central Park’s Bridle Path west of the reservoir in New York City Filmed today, April 15, 2026

@mattsimontacchi has joined our team as President of Field Operations, after spending nearly two decades in field leadership roles at @RedHat. Throughout his tenure at Red Hat, Matt was at the forefront of translating emerging open-source innovation into enterprise adoption. He played a pivotal role in building the field organization and customer engagement models that helped technologies like Linux and Kubernetes evolve from developer-driven projects into foundational elements of global infrastructure. "What I learned at Red Hat is that when an open-source community earns the trust of the world’s best engineering teams, something powerful happens. I saw it with Linux, JBoss, and Kubernetes, and the playbook for bringing that value to the enterprise is changing fast. AI is rewriting how companies build, deploy, and operate software. The opportunity at Astronomer is to take what I know about scaling enterprise field organizations and adapt it for a world where reliable, governed data pipelines are the bridge to the enterprise context that makes agentic AI work at scale. Airflow already has that trust, and now our task is to help organizations turn it into their competitive advantage." – Matt Simontacchi, President of Field Operations, Astronomer Link to the press release: bit.ly/3Q5ggT7

An Osprey stopped by Central Park's Harlem Meer and within two minutes was on its way with a bullhead catfish to go.

🎸 "Astronomer is moving from being a tool for data engineers, to a place where the entire organization will rely on the service to power the data we need." - Miko Chen, Lead Data Engineer, @createmusic Miko spoke with @diginomica about how Create Music Group, one of the fastest-growing names in music tech, uses Astro to orchestrate 600+ data pipelines and deliver real-time insights for music artists globally. 📰 Read the article: bit.ly/3P5vI19

Most text-to-SQL fails at scale. Not because of the LLM — because of context. We built Kepler, our internal AI data assistant, around one insight: the hard problem isn't SQL generation. It's discovery. Context layer over warehouse metadata + Airflow lineage. Hybrid search (keyword + semantic + usage-based boosting via RRF). A discovery subagent that finds the right tables before the main agent writes a line of SQL. Playbooks that institutionalize what worked. A third of Astronomer uses it. 3.5K queries answered in a few months. Powered by the data foundation Astro makes possible. Full writeup on our engineering blog → astronomer.io/blog/building-…

I stood in the rain this morning to capture this tender moment between two ravens nesting at a famous uptown Manhattan cathedral. The female has been on the nest for days and was calling loudly for her mate before he flew in. #birdcpp