Databricks making a big play in Data/AI

First Rays Venture Partners Fund 1 portfolio company Databricks hosted a Data & AI Summit in June, 2024. As partners and investors, we were invited to an exclusive preview of the product and financial performance of the company. First Rays portfolio company, Arcion Labs, which was acquired by Databricks last year was one of the highlights of the conference and was relaunched under the new brand name - ‘Lake Flow’ and will be an important pillar of the data strategy for the company.

The conference saw several significant announcements, marking advancements in data and AI technologies. Here are the key highlights: 

1. Unity Catalog Open Source: Unity Catalog, Databricks’ unified solution for data and AI governance, is now open source. This initiative aims to provide greater flexibility and control for users, promoting an open ecosystem that supports interoperability across various data formats and compute engines.

2. Databricks + Tabular Acquisition: Databricks announced the acquisition of Tabular, a data management company. This move aims to enhance interoperability between Databricks’ Delta Lake and Apache Iceberg, allowing for more seamless data management across platforms.

3. Mosaic AI: Mosaic AI’s capabilities are now fully integrated into the Databricks platform. This includes tools for zero-code model training, agent framework for developing AI systems, and advanced evaluation metrics for AI outputs. The Mosaic AI Tool Catalog allows for the sharing and governance of AI tools within organizations. 

4. Serverless Compute: Databricks announced the general availability of its serverless compute option, emphasizing its future as the platform’s standard. This shift is aimed at optimizing resource usage, cost management, and simplifying deployment.

5. Databricks LakeFlow: A unified data engineering solution that integrates ingestion, transformation, and orchestration of data pipelines on serverless compute, enhancing data quality and operational efficiency. 

6. Databricks AI/BI: A new business intelligence product designed to leverage AI for deeper data insights. It includes AI/BI Dashboards for low-code analytics and AI/BI Genie, a chat-like interface for broader business questions. 

7. Databricks Clean Rooms: This feature allows for privacy-safe collaboration on data and AI projects across different cloud environments, promoting secure data sharing and innovation. 

These announcements reflect Databricks’ ongoing commitment to innovation in data and AI, aiming to provide more open, interoperable, and efficient solutions for their users. During the investor briefing, Databricks also shared that the revenue in GenerativeAI is growing rapidly and the compound AI system through Mosaic platform is generating revenue through AI inference and fine-tuning. The company highlighted how using Compound Systems (RAG + Fine-Tuning) has improved accuracy levels to 89% as against 59% in using commercial LLMs

Our analysis states that Databricks will be a long term winner because of the following reasons:

1. Comprehensive Data & AI platform - with the Delta Lake Uniform that covers all data formats, Delta Lake that can house both structured and unstructured data and Mosaic AI platform, Databricks is now a complete platform to meet the Data & AI needs of Enterprises.

2. Open Source as an Advantage - Databricks has open sourced all parts of the platform including the Unity Catalog. This enables unhindered adoption at the Enterprise level and brings customers to seek value-addition on the platform through Databricks.

3. Expansion from Developer to Business User - As the company develops its AI/BI product, Databricks is expanding its TAM as a Data Intelligence company.

4. Move to 100% Serverless - Databricks announced that it is moving to 100% serverless and will own customers compute as well as storage requirements. This is expected to add significant revenue to the company.
   As the company moves to 100% serverless, Databricks team gave a glimpse of how the revenue of the company will be represented in 2025. By H1‘25, Databricks is expected to have $5 Bn in Revenue with 100% serverless.

In Summary, First Rays team is excited about our investment holdings in Databricks and we believe that the company will demonstrate robust performance and growth in the coming years. We are also engaging many of our portfolio companies in partnership with Databricks and creating greater value to Enterprise Customers.

Databricks Launches Lake Flow (as part of Arcion Labs integration)

DB delivers higher accuracy for enterprise GenAI

DB selling multiple products in its platform capturing Developers & Business Users

The company also shared its strong financial performance for FY ‘24 with over 50% YoY growth