Ahmet Altay

We are wrapping up the #BeamSummit with the 3rd #RAG talk today (I guess it is a popular thing😉) and CI/CD & cost optimizations for #Dataflow. Lots of insights on running @ApacheBeam and #Dataflow was shared. If you missed it, all talks will be on YouTube in a few weeks.

@janl_apache @ml6team @ApacheBeam I "shorten" things to reduce into a tweet format but otherwise the talk addressed the why. All talk videos will be available on YouTube in a few weeks.

@iridium77 @ml6team @ApacheBeam Yes! But please share the most important question. Why?

Quote from @ml6team at the #BeamSummit "@ApacheBeam is super useful, good for ML, super flexible, a prime candidate for building these things". These = {RAG, multi-modal, embedding, vector DBS}

Simultaneously at the #BeamSummit Namita Sharma and Pablo Rodriguez Defino are showing a ready to use RAG architecture complete with a GitHub repo. beam.apache.org/blog/dyi-conte…

Processing 6 million events per minute

Why Beam? Data scientists at Lyft prefer Beam python over Flink python API.

"How @ApacheBeam has helped us significantly to build our realtime forecasting platform at @lyfteng" -- Ravi Magham at the #BeamSummit

From the speakers: "Beam YAML made our life a lot easier!" 🤗

Nate Salawe followed up with a live demo, showing the deployment happening in less than 2 minutes! ⚡

@cadetiloye from @mavencode shared how BeamStack (built in top of Beam YAML) could be quickly deployed on any public cloud or on prem and makes LLM, embedding, vector DB workloads easy and observable. #BeamSummit

Debugging #Dataflow and @ApacheBeam ML? Easy. Rajkumar Gupta is covering the common issues at #BeamSummit.

#BeamSummit is starting the second day with Uday Kalra's keynote on how Google DeepMind is using @ApacheBeam for large scale & accelerated ML inference.

Day 1 of #BeamSummit 2024 sounds incredible! Loved hearing about all the innovation in AI and data platforms, especially the cool use cases from @Cruise, @Uber, and Project Shield. Excited to see where #BeamML goes next! Can’t wait for more tomorrow! #BeamSummit2024 #Innovation @ApacheBeam

Lydian Lee from @Affirm wrapped up the first day of #BeamSunmit with how they built a robust internal platform using @ApacheBeam . See you all at the networking hour today and tomorrow at the day 2 of the summit!

Olu Akinlaja from @doitint long time @ApacheBeam and #Dataflow is talking about using dynamic destinations with cloud #pubsub at the #BeamSummit.

ML experts from @ml6team showing multi-modal LLM use cases with real-world examples using @ApacheBeam at the #BeamSummit.

We are continuing the #BeamSummit in the afternoon with how @ApacheBeam efficiently serves large ML models & how this is all made easy to use with YAML and #Dataflow job builder (UI only streaming ML and analytics)

"We migrated ~5 TB (~2 billion rows) of production NoSQL DB to a Postgres DB without application downtime." @ApacheBeam is used at Recursion (a Biotech company) to continue operations and drug discovery pipeline with ease and without interruption. #BeamSummit

We are welcoming @UberEng for the first time at the @ApacheBeam #BeamSummit covering the Michelangelo platforms batch prediction capabilities!

"@ApacheBeam reduced our costs by 90%"

#BeamSummit continues with a keynote from Marc Howard on how Project Shield defends journalists, civil organizations, human rights organizations at most critical times powered by @ApacheBeam.