StarTree Unveils New AI-Native Features to Boost Real-Time Analytics

Shutterstock

Big data startup StarTree unveiled new enhancements to its platform on Wednesday, designed to improve AI-driven decision-making for enterprises.

The company has launched AI-native features, including Model Context Protocol (MCP) and Vector Auto Embedding, while also making its Bring Your Own Kubernetes (BYOK) deployment model generally available.

StarTree is known for its real-time analytics platform built on Apache Pinot – an open-source, distributed OLAP data store optimized for ultra-low-latency queries. Its technology enables enterprises to process vast streams of dynamic data, delivering AI-powered insights, anomaly detection, and interactive dashboards to support real-time operational decision-making.

As AI adoption accelerates, traditional analytic systems designed for human review are proving too slow. Businesses increasingly require AI-native architectures capable of handling real-time data streams for autonomous decision-making.

The introduction of the MCP provides a standardized way for AI applications to interact with external live data sources and tools, ensuring that decisions are made with the most current context available. Originally developed by Anthropic to help applications provide context to LLMs, MCP allows these models to pull real-time insights from StarTree, giving them the ability to act beyond their original training. The feature is expected to launch in June 2025.

While the MCP feature is primarily aimed at facilitating developers, it also has the potential to improve AI-driven applications by making integrations more scalable, efficient, and adaptable across various platforms. A key benefit of MCP is that it adds some muscle to conversational querying in real-time. This means that users can ask follow-up questions without restating the full context, and the AI can still deliver accurate, real-time answers

The other new feature, the Vector Auto Embedding, is designed to optimize the transformation of vast streams of raw, dynamic data into actionable intelligence, simplifying real-time retrieval-augmented generation (RAG). The feature enables pluggable vector embedding models, streamlining the continuous flow of data from source to embedding creation to ingestion. 

According to StarTree, this approach makes deploying Retrieval-Augmented Generation pipelines easier by reducing complexity and eliminating stitched-together workflows. The company also claims that these innovations minimize latency and enhance the responsiveness of autonomous agents in high-speed environments. Vector Auto Embedding is expected to be available by Fall 2025.

These combined capabilities enable StarTree to power agentic AI applications, facilitate RAG, and support natural language querying of live data with immediate context. 

“The next wave of AI innovation will be driven by real-time context—understanding what’s happening now,” said Kishore Gopalakrishna, Co-founder and CEO of StarTree. “StarTree’s heritage as a real-time analytics foundation perfectly complements where AI is going by delivering fresh insights at scale. What is changing is the shift from apps as the consumer to autonomous agents.”

Shutterstock

The new Bring Your Own Kubernetes (BYOK) offers greater control to developers over StarTree within Kubernetes environments, regardless of whether it is in the cloud, on-premises, or hybrid architectures.

This feature is especially useful for companies that handle sensitive data, as it allows them to meet strict security and compliance requirements within their own tech environment. It’s also a cost-effective solution for companies with predictable and stable workloads, helping them reduce expenses on compute and usage fees.

“Real-time insights are no longer optional, but too often, enterprises are blocked by infrastructure constraints,” said Gopalakrishna. “With BYOK, we remove those barriers. Companies can now deploy StarTree wherever they need it, without compromising on performance, security, or cost control.”

The BYOK feature, now available in preview, adds more flexibility to StarTree’s offerings. The startup already provides a fully managed SaaS service, where StarTree handles hosting for users, as well as the Bring Your Own Cloud (BYOC) option, which lets users select their preferred cloud while StarTree manages the deployment. Now with BYOK, users can host and manage everything themselves.  

StarTree, headquartered in Mountain View, CA, was founded in 2018 by Kishore Gopalakrishna and Xiang Fu. Both the founders were involved in the development of Apache Pinot while working at LinkedIn. Frustrated by the inefficiencies in real-time analytics that slowed workflows, Gopalakrishna and Fu set out to create a purpose-built platform that could deliver real-time, high-concurrency analytics at scale. 

Their vision led to the founding of StarTree, which has raised a total of around $ 75 million in funding. The growth-phase company claims its platform ingests 1.1 million events per second while serving 1.19 billion queries per week.  It helps businesses like Stripe, Uber, DoorDash, and Zomato in delivering real-time, user-facing analytics at scale. 

Related Items 

Unpacking How Enterprises Can Source and Utilize Synthetic Data for AI

Event Store Changes Name to Kurrent, Raises $12M to Unify Streams and Databases

Demystifying AI: What Every Business Leader Needs to Know