The data catalog that grew up into the trust layer for enterprise AI. Built in Redwood City on one stubborn idea: people will only use data they can actually trust.
Somewhere inside a Fortune 100 bank, an analyst types a plain-English question about last quarter's exposure. They never touch a query console. An Alation agent reads the request, finds the right table, checks who certified it, confirms the access policy, and answers - with the lineage attached.
That quiet exchange is the whole point of Alation. The company spent more than a decade building a map of where enterprise data lives and whether anyone should believe it. Now it is sending software agents to walk that map on your behalf. The catalog, once a glorified inventory list, has started to act.
Alation calls the current version an "Agentic Data Intelligence Platform." Strip the jargon and the job is older and simpler: connect the people with questions to the data with answers, and make sure the answer can be trusted. It is unglamorous work. It is also the bottleneck standing between most companies and the AI they keep promising shareholders.
By the early 2010s, enterprises were drowning. Warehouses multiplied, dashboards bred dashboards, and somewhere in the sprawl sat the one table that actually mattered - next to nine that looked exactly like it and lied. Asking "which number is right?" usually meant pinging a colleague who had since changed teams.
The tools of the day treated this as a storage problem. Alation's founders treated it as a knowledge problem. Data was not missing; the context was. Who made this table? Who uses it? Is it stale? Should you even be allowed to see it? Answers existed - scattered across human heads and forgotten wikis - just nowhere you could search.
On December 12, 2012, four people incorporated Alation in Redwood City: Satyen Sangani, Aaron Kalb, Venky Ganti, and Feng Niu. Sangani, who had spent years at Oracle, had grown tired of databases that demanded a priesthood to interpret them. He and Kalb kept circling the same question - how do you connect people who have questions to people who have answers?
Their bet was a hybrid. Let machine learning crawl the logs and surface what people actually query, then let humans curate and certify the results. Pure automation missed nuance; pure manual cataloging never finished. The combination became the first product - a natural-language way to ask a database questions, years before that idea got fashionable - and helped define an entire category: the modern data catalog.
Co-founder & CEO. Ex-Oracle. Host of the Data Radicals podcast. The economist who decided data needed a friendlier front door.
Co-founder. The human-computer-interaction half of the original question about questions and answers.
Co-founder. Brought the machine-learning depth that let the catalog read behavior, not just schemas.
Co-founder. Helped turn academic data-systems research into shipping enterprise software.
Alation incorporates and ships a natural-language data querying tool that helps define the modern data catalog.
Costanoa Ventures leads early funding to widen the product and the customer base.
A ~$50M Series C arrives as data governance and self-service analytics become board-level concerns.
A $110M Series D led by Riverwood Capital - with Snowflake Ventures aboard - values Alation at $1.2B.
Thoma Bravo leads at a $1.7B+ valuation; Databricks Ventures joins, putting both data-cloud giants on the cap table.
Alation launches its Agentic Platform and AI Agent SDK, then acquires Numbers Station AI to power agentic data workflows.
What started as a search box is now a portfolio. The Data Catalog unifies discovery with natural-language search, surfacing definitions, lineage, policies, usage, and trust signals - kept current by active metadata and more than 120 connectors. Around it sit modules for governance, quality, and lineage, each solving a piece of the same trust puzzle.
The cleverest move is meeting people where they already work. Alation Anywhere drops trusted data into Excel, Teams, and Slack, so nobody has to learn a new app to act on a certified number. And the newest layer - Chat with Your Data and a growing roster of agents - lets users interrogate structured data in plain language instead of code.
Natural-language discovery with lineage, policies, usage, and trust signals - fed by active metadata and 120+ connectors.
Centralized policies plus automated stewardship, access, masking, and approvals across the estate.
Quality metrics shown right in the catalog, via an Open Data Quality Framework that plugs into best-of-breed tools.
Trusted data delivered inside Excel, Teams, and Slack - the apps people actually live in.
AI agents automate discovery and governance; an SDK lets customers build their own data agents.
Ask questions of structured data in natural language - no SQL required.
Adoption is the only review that counts in enterprise software, and Alation's is broad: roughly 450 customers across finance, healthcare, pharma, manufacturing, retail, insurance, and tech, including more than a quarter of the Fortune 100. Revenue tells a similar story - the company crossed $100M in annual recurring revenue in September 2022 and reported about $109M in revenue for 2024.
The investor roster is its own kind of proof. Snowflake and Databricks compete fiercely for the same data workloads - yet both put money into Alation, betting that whoever wins the warehouse war, somebody still has to catalog the spoils. Thoma Bravo, Riverwood Capital, Costanoa Ventures, and Sanabil Investments round out the ~$340M raised.
Alation frames its purpose as empowering a curious and rational world - making it easy for people to find, understand, trust, and use data. It sounds lofty until you remember the alternative: decisions made on the wrong table, and AI models trained on data nobody vetted.
That is the tension the company keeps living inside. Governance, done badly, is a brake that slows everyone down. Done well, it is the seatbelt that lets people drive faster. Alation's whole pitch is that trust and speed are not opposites - that the catalog can be the thing that says "yes, use this" instead of the committee that says "wait."
Every enterprise now wants AI that answers questions about its own business. The dirty secret is that those answers are only as good as the data underneath - and most companies cannot say with confidence which data is right. That is precisely the gap Alation has spent fourteen years mapping.
Its 2025 turn toward agents pushes the bet further: instead of humans clicking through a catalog, agents will read intent, fetch trusted data, and operate inside the guardrails automatically. The competition is real - Collibra, Atlan, Informatica, Microsoft Purview - and the category Alation helped invent is now crowded. But being the neutral map in a contested territory is not a bad place to stand.
Return to that bank analyst from the opening. A decade ago, their question would have meant a Slack thread, a stale wiki, and a guess. Today an agent answers it, with the lineage attached and the policy already checked. The data did not change. What changed is that someone finally made it trustworthy enough to ask.
Editor's note: Figures (valuation, revenue, customer counts, funding) are drawn from public reporting - Crunchbase, TechCrunch, CB Insights, GetLatka, and Alation's own announcements - and are approximate. Private companies rarely confirm exact numbers; treat them as well-sourced estimates rather than audited fact.