Meta has acquired Sundial, an AI-powered data analytics startup co-founded in 2021 by two of its own former executives, Julie Zhuo and Chandra Narayanan, a company spokesperson confirmed to Axios on October 23. Meta did not disclose the price. The startup’s technology and staff will be folded into Meta’s new data analytics organization, and Sundial will stop serving its external customers once the deal closes — a notable list that includes OpenAI, one of Meta’s fiercest AI rivals. Narayanan, Meta’s former analytics chief, will lead the company’s data analytics tooling and agents work under chief data officer Alex Schultz.
Two Alumni Return With Their Startup
Zhuo and Narayanan are among the most prominent Meta alumni to build startups and come back. Zhuo was Meta’s vice president of product design; Narayanan led analytics at Sequoia Capital and Instagram and served as Facebook’s director of analytics before leaving to found Sundial. According to reporting on the deal, several other Sundial employees are also Meta alumni, making the acquisition as much a homecoming as a technology purchase.
Sundial’s pitch was self-service analytics for enterprises: tools that turn natural-language questions and complex datasets into actionable insights without routing every query through a central data team. The company combined AI, contextual data and what it called expert playbooks to help teams make decisions faster. It raised $23 million in total funding, including a $16 million Series A in July 2025 led by DJ Patil of GPV, with backing from Sequoia Capital.
Where Sundial Lands Inside Meta
The deal will not create a new product line. Sundial’s tools and team are being absorbed into Meta’s Data Analytics Tooling & Agents unit — nicknamed DATA — which Schultz was appointed to lead when Meta named him its first chief data officer in July 2026. The unit handles descriptive analysis, experimentation and monitoring of Meta’s critical AI systems, issuing alerts when those systems misbehave.
“AI has created effectively unlimited demand for insights, and we want to define the future of AI analytics,” Schultz wrote in a LinkedIn post announcing the acquisition, which also confirmed Narayanan’s new role leading the analytics tooling and agents work. In April 2026, Sundial joined the Open Semantic Interchange, an initiative led by Snowflake to standardize how data metrics are defined across AI tools — a sign of how closely the startup had been tracking the industry’s agent-driven analytics wave.
The OpenAI Problem
The awkward detail in the deal is Sundial’s client list. OpenAI, Meta’s direct competitor across AI models and products, is one of Sundial’s customers — and Sundial plans to end external service when the acquisition closes. Neither company has said what happens to those customer relationships, or whether OpenAI’s usage data and analytics workflows migrate with the rest of the business. For a company racing Meta in foundation models, continuing to run analytics through a Meta-owned platform would be an unusual arrangement.
The acquisition also fits the broader industry pattern of large technology companies buying specialized AI analytics capabilities rather than building them. With no price disclosed and external service ending, Sundial reads as a targeted talent-and-tools deal: Meta gets Narayanan and Zhuo back, absorbs a proven analytics stack, and removes one more independent option from the market.
Why It Matters
Sundial’s real significance for Meta is internal. The company is spending $135 billion to $145 billion on AI capital expenditures this year while free cash flow has turned negative, and running that buildout requires analytics infrastructure that can keep up with thousands of experiments and agentic workflows. Schultz’s DATA unit was created precisely to answer questions Meta hasn’t even thought to ask yet, and Sundial’s self-service tools were built for exactly that job.
For Meta’s broader AI strategy, the deal underscores how much of the current race is about data plumbing, not just models. The company’s assistant ambitions depend on understanding what users and systems do at scale — and with Sundial’s founders back in the building, Meta now controls one more piece of that machinery outright.
Background
Meta’s AI spending spree has made acquisitions a recurring part of its strategy: the company has been buying talent and tooling across the AI stack as it builds out its models and the infrastructure behind them. Sundial, founded in 2021 by two executives who helped shape Meta’s own product and analytics culture, spent five years building self-service analytics tools for enterprise customers before returning to its roots. The deal also arrives days after Meta launched its Muse personal agent in Japan, as the company pushes its AI products into new markets.
0 Comments