Data + AI Summit 2024
2024-06-10..2024-06-13
-- online --
| title | author | length |
|---|---|---|
| Keynote Day 1 | Ali Ghodsi | 28m |
Keynote lays out three concrete enterprise GenAI problems: 85% of GenAI pilots not in production, fragmented data formats and lock‑in, and security/privacy… | ||
| Keynote Day 1 - Full | Ali Ghodsi, Brian Ames, Patrick Wendall, Jackie Brosamer, Fei Fei Li, Jensen Huang, Reynold Xin, Ken Wong | 02h 52m |
Databricks unveils a concrete stack: Databricks Data Intelligence Platform with full serverless availability July 1, native NVIDIA GPU acceleration, and DBRX… | ||
| On Building Production-Quality AI Systems | Patrick Wendell | 37m |
Databricks announces specific production features: zero-code fine-tuning for open-source models, a hosted vector index with state-of-the-art embeddings, a… | ||
| The Best Data Warehouse is a Lakehouse | Reynold Xin, Pearl Ubaru | 23m |
Databricks SQL claims concrete gains: over 7,000 customers on a Lakehouse built on Delta Lake and governed by Unity Catalog. Cold start time for a warehouse… | ||
| Jensen Huang, Founder and CEO of NVIDIA with Ali Ghodsi, Co-founder and CEO of Databricks | Jensen Huang, Ali Ghodsi | 25m |
NVIDIA and Databricks announce a partnership to run GPU-accelerated data processing for Databricks’ Photon engine, shipping a suite of optimized libraries… | ||
| On the History and Future of AI | Fei Fei Li | 18m |
Tells how modern vision AI combines neural networks, GPUs and a curated ImageNet of 50 million images to move from labeling to generative and 3D models; cites… | ||
| On Bringing AI to Business Intelligence | Ken Wong | 17m |
Databricks launches Databricks aibi, a BI layer plus a conversational agent called Genie that runs inside Databricks SQL and uses Unity Catalog metadata. Demo… | ||
| Building an Insights Factory | Brian Ames | 07m |
GM builds an Insight Factory on Databricks with a React UI, Medallion tables, Unity Catalog, Amuta for policy, Atlan for lineage and MLflow for production… | ||
| Building and Deploying GenAI Apps at Block | Jackie Brosamer | 07m |
Block builds a federated GenAI stack on top of Databricks AI Platform using Mosaic AI Gateway, MLflow model ops, model serving and Unity Catalog for… | ||
| On Data Intelligence | Ali Ghodsi | 01m |
Customers want custom AI on your data while keeping cost and privacy under control, and the talk frames that problem as data intelligence. Mosaic AI is… | ||
| Keynote Day 2 - Full | Alexander Booth, Ali Ghodsi, Bilal Aslam, Darshana Sivakumar, Hannes Mühleisen, Matei Zaharia, Reynold Xin, Ryan Blue, Tareef Kawaf, Yejin Choi, Zeashan Pappa | 02h 16m |
Announces include Uniform GA with 4+ exabytes loaded and Delta processing cited at ~9 exabytes, plus a 73% BI performance uplift over two years; DuckDB adds… | ||
| Closes the Keynote with a Summary of Product Announcements | Ali Ghodsi | 01m |
Genie and aibi now expose open APIs so BI tools can query models directly and the platform claims improved data intelligence for natural language on your data… | ||
| Day 2 Opening Remarks | Ali Ghodsi | 03m |
Announces the acquisition of Taber and the GA of Uniform (Universal format) that adds interoperability between Delta Lake and Apache Iceberg so a single store… | ||
| How to Make Small Language Models Work | Yejin Choi | 18m |
Starts from a tiny GPT-2-style teacher and overgenerates outputs, uses aggressive three-stage filtering (including a PMI-style information-theoretic filter) to… | ||
| Unity Catalog Demo of New Features | Zeashan Pappa | 05m |
Live demo shows a Snowflake table federated via Lakehouse Federation joined with a native Lakehouse returns table and turned into a Unity Catalog managed… | ||
| The Future of Lakehouse Format Interoperability | Ali Ghodsi, Ryan Blue | 05m |
Databricks acquires work on Delta Lake and commits to making it interoperable with Apache Iceberg, aiming to eliminate format choice by standardizing table… | ||
| Evolving Data Governance With Unity Catalog | Matei Zaharia | 15m |
Announces an open-source Unity Catalog (v0.1) now public on GitHub and accepted by the Linux Foundation, ships table, volume (unstructured) and AI… | ||
| Lakehouse Format Interoperability With UniForm | Shant Hovsepian | 06m |
UniForm ships as GA tooling that lets you write Delta and read as Iceberg/Hoodie by translating metadata to other open lakehouse formats with minimal… | ||
| Announcing Databricks Clean Rooms with Live Demo | Matei Zaharia, Darshana Sivakumar | 08m |
Databricks launches a clean rooms product as Public Preview on AWS + Azure that runs arbitrary computations including Python ML code and notebooks not just… | ||
| Data Sharing and Cross-Organization Collaboration | Matei Zaharia | 06m |
An open protocol called Delta Sharing serves Parquet files so organizations can publish parts of tables directly to pandas, Apache Spark, Tableau, Power BI or… | ||
| The Evolution of Apache Spark™ | Reynold Xin | 11m |
Apache Spark now treats Python as a first-class language with API, error-message and performance work; PySpark downloads span 200+ countries, and Databricks… | ||
| How Data Intelligence is Delivering Big Wins at Texas Rangers | Alexander Booth | 11m |
Team uses Databricks Lakehouse with Unity Catalog, ML registry and a feature store to unify sensitive biomechanical, medical and scouting data and enforce… | ||
| Code-First Data Science | Tareef Kawaf | 05m |
Posit is a 15‑year public benefit corporation that pushes a code‑first approach, maintaining hundreds of open‑source packages and the RStudio IDE while adding… | ||
| Databricks LakeFlow: A Unified, Intelligent Solution for Data Engineering | Bilal Aslam | 17m |
Databricks announces LakeFlow as a unified product that ingests from Salesforce, SQL Server and apps via Archon-powered native connectors, transforms with… | ||
| Preview of Apache Spark 4.0 and GA of Spark Connect | Reynold Xin | 02m |
Spark Connect goes GA and becomes the default client protocol in Spark 4.0 while NC SQL is slated to be the standard SQL surface; the project team publishes a… | ||
| Recap of Announcements | Ali Ghodsi | 01m |
Recap lists concrete product moves: Unity Catalog open source is demoed live, a real open-source release rather than a roadmap bullet; a new metrics capability… | ||
| Announcing DuckDB Support for Delta Lake and a DuckDB Extension to Unity Catalog | Hannes Mühleisen | 05m |
Announces DuckDB 1.0 with a stable SQL dialect and a backwards‑compatible storage format, plus official Delta Lake support that reads Delta tables with zero… | ||
| Open Sourcing Unity Catalog Live Onstage | Matei Zaharia | 01m |
Unity Catalog becomes an open source project live onstage: a GitHub repository is created, navigated to the settings, and toggled from private to public using… | ||
| Announcing Delta Lake 4.0 with Liquid Clustering | Shant Hovsepian | 05m |
Delta Lake 4.0 ships a new layout and type system: Liquid Clustering replaces manual partitioning and reports up to 7x faster writes and 12x faster reads on… | ||
| The Evolution of Delta Lake | Shant Hovsepian | 16m |
Announces Delta Lake UniForm GA — a universal format that reads Delta as Iceberg or Hudi with minimal overhead, already moving over four exabytes of data and… | ||
| What's Next for Apache Spark™ Including the Upcoming Release of Apache Spark 4.0 | Reynold Xin, Tareef Kawaf | 19m |
Preview of Spark 4.0 shows Spark Connect becoming GA as the language‑agnostic gRPC + Apache Arrow binding that decouples driver and server, enabling per‑client… | ||
| Databricks Customers | 01m | |
Over 250 customer stories summarize concrete outcomes: a sports company deploys a predictive model that forecasts future player performance and feeds lineups… | ||
| Databricks Product Announcements | 05m | |
Announces that Databricks makes the platform serverless everywhere, adds zero‑code fine‑tuning for open‑source models and hosts Shutterstock’s new image model… | ||
| nothing matches | ||