Partner Story · in conversation with Sundial. This is a Partner Story — a written interview with a company building at the center of AI, published through our free Submit Your Story feature. The claims below are Sundial's own, shared in their words and edited for clarity. No payment was involved. (Where we could, we checked the big ones: the funding, investors, and OpenAI relationship are all a matter of public record.)
Most companies don't have a data problem. They have a decision problem. The data is there — usually far too much of it — but the answer a business actually needs is trapped behind days of SQL, a queue of analyst requests, and five dashboards that don't agree with each other. By the time the number lands, the decision it was supposed to inform has already been made without it.
That gap — call it decision latency — is what Sundial is built to close. And the money math is blunt: an analyst's time is expensive, and a slow answer to "why did retention drop last week?" can cost far more than the salary of the person who eventually finds it. Here is Sundial's own account of the platform that some of the most data-sophisticated companies in the world now run on.
Sundial is an AI-powered analytics platform for B2B companies that need fast, trustworthy answers from their data. It was co-founded by Chandra Narayanan and Julie Zhuo — former analytics and product leaders from Meta, Instagram, and Sequoia Capital. Zhuo was VP of Product Design at Facebook and is the author of The Making of a Manager; Narayanan was Chief Data Scientist at Sequoia. The company is headquartered in Menlo Park, with a team spanning the US and India.
Where does AI actually sit in the product?
"AI isn't a layer on top of Sundial. It's the engine," the team told us. Instead of a human analyst writing queries and assembling reports, Sundial's agents do that work autonomously — an agentic workflow that replaces the traditional analytics pipeline.
But the part they're most insistent about is what keeps the AI honest. "What separates us from a generic LLM-over-data tool is that our agents operate on top of expert-built playbooks — analytical frameworks written by people who spent careers doing growth analytics at companies like Meta, Airbnb, and Sequoia. The AI executes with speed; the playbooks ensure it executes with rigor." Every answer, they say, ships with its source evidence attached, so a user can see what a conclusion is built on rather than trusting a number on faith.
The analytics market is crowded. What's the actual difference?
The pitch is a sharp one: most BI tools give you more charts to stare at; Sundial says it gives you decisions. "A product manager shouldn't need a data science degree to understand why retention dropped last week. An executive shouldn't have to wait three days for an analyst to run a cohort analysis." The bet is that turning both of those into something instant — and grounded in evidence — is worth more to a business than another dashboard.
Proof and traction — who's actually using this?
Sundial counts OpenAI, Gamma, Character, and Mighty Networks among its customers. In a published case study, OpenAI's VP of Analytics, David Sasaki, described going "from raw logging to comprehensive data visibility virtually overnight," crediting Sundial with automating data-engineering work that would have taken a small internal team months to build.
The funding backs it up. Sundial has raised $23 million in total, including a $16 million Series A led by DJ Patil — the first-ever US Chief Data Scientist — with Sequoia Capital, Slow Ventures, and Tribe Capital among the investors. For a company selling rigor in AI analytics, having the person who literally coined the modern "data scientist" role lead your round is not a bad signal.
What was the hardest lesson in building it?
By Sundial's own account, the sharpest lesson came from an early failure. "AI analytics tools fail not because the AI is bad," the company says, "but because the AI doesn't know what 'good analysis' looks like for your business."
When they first pointed the model at customer data, it produced answers that were "statistically accurate but analytically naive" — computing the average where you need the median, or looking at 7-day windows when the business actually runs on 28-day cycles. The fix wasn't a better model. It was baking human expertise into the system rather than bolting it on afterward: former data-science leaders from Meta and Airbnb now spend the first 30 days with each new customer setting up metrics, context, and playbooks. "Speed doesn't matter if the answer is wrong."
It's a lesson that generalizes well beyond analytics — and one worth remembering for anyone building an AI product on top of a foundation model right now.
What's next?
Sundial is moving from reactive analysis — ask a question, get an answer — toward what it calls proactive intelligence, where the system surfaces the insights that matter before anyone thinks to ask. It's also building out a self-serve tier so teams can start without a long enterprise procurement cycle.
If you're a B2B company sitting on data that isn't driving decisions fast enough, you can learn more or request access at sundial.ai.
---
Partner Stories are written interviews with companies building in AI, published through our free Submit Your Story program. The statements above are the company's own and reflect its perspective; they are not verified claims or endorsements by AI Business, and no payment was exchanged for publication. Want your company featured? Submit your story.
