Meta has acquired Sundial, an AI-powered data analytics startup co-founded by two of the company's best-known former executives, Axios reported exclusively on Friday, in a deal designed to pull veteran talent back inside the company and reshape how Meta's own employees interrogate its mountains of data.
The price was not disclosed. Sundial raised $23 million as an independent company, including a $16 million Series A in July 2025 led by DJ Patil of GPV, making this a small acquisition by Meta's standards — but one with outsized internal significance. For more context on this story, see our ongoing breaking AI news.
The Founders Are Coming Back
Sundial was co-founded in 2021 by Julie Zhuo, Meta's former vice president of product design and one of the most recognizable product leaders of the social media era, and Chandra Narayanan, previously analytics chief at Instagram and later at Sequoia Capital. Both are now returning to Meta with members of their team, according to the Axios report.
Narayanan will lead data analytics tooling and agents work inside a newly formed unit nicknamed DATA, short for Data Analytics Tooling & Agents. The group reports to Alex Schultz, the longtime growth executive whom Meta named its first-ever chief data officer in July.
"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 deal.
The message from Meta is that business intelligence is becoming a conversation. Sundial's platform turned natural-language questions into rigorous analysis through a chatbot plugin, letting users query complex datasets the way they would ask a colleague — an approach that fits squarely with the agent-driven direction of Meta's internal tooling.
What Meta Gets
Sundial's customer list included OpenAI, a detail that gives the small startup an outsized credential: one frontier lab was already paying to make data analysis conversational. Once the deal closes, Sundial will stop serving external customers entirely, folding its product into Meta's internal stack.
For Meta, the acquisition is less about revenue than about capability. The company sits on some of the largest data collections in the private sector, spanning its family of apps and its advertising engine. Turning that data into answers has traditionally required analysts, dashboards, and SQL. The bet behind DATA is that natural-language agents can collapse that pipeline, putting analysis directly in the hands of product teams, executives, and advertisers.
A Small Deal in an Expensive Quarter
The purchase lands at a delicate financial moment. Meta's third-quarter results, reported this week, showed capital spending of $34.9 billion exceeded operating cash flow of $32.8 billion, pushing quarterly free cash flow to negative $2.1 billion. Over the first nine months of 2026, free cash flow fell 63 percent to $11 billion as AI infrastructure spending surged, and the company continues to guide 2026 capital expenditures to $135 billion to $145 billion.
Against that backdrop, buying a $23-million-funded analytics startup is a rounding error. But the deal illustrates where Meta believes its AI money should flow: not just into training compute, but into tools that make the resulting systems useful to the employees and advertisers who generate the cash flow that funds it all.
The timing also matters internally. Meta told investors this week that its AI investments are already delivering substantial incremental advertising revenue through stronger engagement, better recommendations, and more efficient monetization. An analytics unit that shortens the distance between a question and a defensible answer is, in effect, infrastructure for that same advertising engine — helping the people who tune it act on what the models find.
Data Becomes an Agent Battleground
The formation of a dedicated Data Analytics Tooling & Agents unit — and the installation of a chief data officer above it — signals that Meta sees agentic analytics as a distinct discipline rather than a feature bolted onto existing BI software.
The broader context is a land grab for the future of how companies query data. Startups and incumbents alike are racing to replace the dashboard with the dialogue, and every large AI lab is building analysis agents of its own. Meta's move is notable less for the size of the check than for the strategy: rather than buying a product for customers, it is buying a team to serve itself first.
There is a certain symmetry, too. A startup whose customers included OpenAI now belongs to one of OpenAI's fiercest competitors, and its founders have rejoined the company they left to build it. In an industry famously bad at retaining veteran talent, Meta has found that the fastest way to get its alumni back is to buy what they built.
For Sundial's team, the trade-off is scope versus independence: the platform's external customer relationships end when the deal closes, but the technology now gets to operate on one of the richest data collections in the private sector, with the backing of a company spending at heights almost no startup can match.
---
Stay Ahead of AIThe AI business landscape shifts daily. Get the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
