Starburst is making big moves in the world of AI and data management. These updates are designed to help companies build smarter workflows with multiple AI agents and keep better control over their usage. They also improve how users access and search through different types of data stored in various systems.

Building Smarter AI Workflows with New Tools

Starburst recently released new AI services to its platform. These include the Starburst AI Agent and AI Workflows, which are now part of the Starburst Enterprise Platform and Starburst Galaxy. The goal is to help companies create multi-agent workflows—where several AI tools work together—and monitor how they’re used. This makes it easier to run complex AI tasks without a lot of fuss.

To support this, Starburst added a new MCP server and agent API. These tools help companies connect their AI agents smoothly. Many industry players are racing to develop MCP servers because they make it easier for businesses to set up multi-agent systems. For example, Databricks offers MCP support through managed servers and custom setups, while Snowflake has a managed MCP server in preview and open-source projects.

Better Control and Cost Management for AI Agents

As AI agents become more common, managing them is a big concern. Starburst is rolling out dashboards that let teams keep an eye on how much these agents are used, helping to track costs and ensure compliance. This kind of governance is crucial because AI agents can act independently, chaining tools and data together without direct human oversight.

Experts say that controlling what AI agents do and understanding their actions is key. Stephanie Walter from HyperFRAME Research notes that companies need ways to trace why an agent made certain decisions and to set limits or turn off agents if needed. She adds that as more AI agents are deployed, managing them safely and ethically will be a major challenge. Having good governance tools will set companies apart.

Starburst’s rivals, like Databricks and Snowflake, already offer governance features. Databricks provides tools like the Mosaic AI Gateway, which includes rate limits, usage logs, and cost tracking. Snowflake’s Cortex AI Observability offers similar features, including tracing and cost analysis, along with community-built dashboards. For companies, these features aren’t just helpful—they’re becoming essential.

Unified Access to Multiple Vector Data Stores

Another exciting update from Starburst is its support for vector search and unified access to different vector stores. This means users can perform advanced retrieval tasks across systems like Iceberg, pgvector, and Elasticsearch. The ability to search efficiently across multiple data stores is important for tasks like retrieval-augmented generation (RAG) and complex searches.

While vector search technology isn’t new, what makes Starburst stand out is its ability to connect to multiple vector data sources seamlessly. This gives users more flexibility and power when working with large datasets. These features are expected to roll out by the end of the year, though Starburst hasn’t announced specific prices yet.

In summary, Starburst’s latest updates aim to make AI workflows more intelligent, manageable, and versatile. For businesses, these tools can mean better control over costs, improved compliance, and more powerful data retrieval options. As AI continues to evolve, platforms like Starburst are helping companies stay ahead by offering smarter, more integrated solutions.

Inspired by

• https://www.infoworld.com/article/4070255/starburst-pushes-lakehouse-boundaries-with-multi-agent-ai-and-unified-vector-search.html

Sources

• gmpg.org
• starburst.io
• databricks.com
• databricks.com
• podigee.com