.png)
When we first showed Enigma to the world in 2013, we had seen what was possible when small details got stitched together from data already out in the world. Search for a company and discover datasets you didn't know existed. Stumble onto something unexpected that revealed how the business actually operated. See the detailed, multifaceted activities of a business or the behaviors of the whole economy, all from data that already existed but needed to be connected to be fully understood.
That was the original promise. Take the scattered filings and registrations and contracts that governments publish, connect them, and let people discover things they didn't know to look for.
Over the years, Enigma tuned our data and our architecture to focus on specific problems. We built KYB verification that financial institutions relied on. We assembled sales and marketing data that helped companies find undiscovered prospects. These were real products that solved real problems. But somewhere along the way, we lost something in our core products. The ability to see the connected picture. The unexpected discovery and the capability to see exactly what a business was doing. The moment where you click through three datasets and suddenly understand something about how a business actually operates.
By the end of 2024, we knew we had to make a choice. Keep building on top of an architecture that made each new thing harder than the last, or go back to the foundation and rebuild.
graph-model-1 debuted in March. It's a new data architecture that represents how businesses actually exist: a brand identity that customers see, a legal entity that files paperwork, physical locations where things happen, and the connections between all of them. Your corner coffee shop has a name on the awning, an LLC on file with the state, and maybe a second location across town under the same ownership. These are all the same business. Our data model now knows that.
When we publicly launched graph-model-1 in July for all users, the new Enigma Console launched alongside it. You can build lists using semantic search, the way you'd actually describe what you're looking for rather than hunting through filters. You can upload a file and get it enriched with revenue signals, industry classifications, ownership data. And when you pull up a business, you see the connected picture again. The legal entity, the DBAs, the locations, the economic activity. What we could show in a demo twelve years ago, but now at scale, and extraordinarily accurate.
The KYB API got rebuilt on top of this new foundation. More verifications resolve automatically. Fewer get kicked to manual review. The data is fresher because the underlying system was designed for freshness from the start, not patched to support it.
Those were the products we already had, made better. But the rebuild also made new things possible.
We built the Enigma AI Connector, a way for AI models to query Enigma directly using MCP: from graph-model-1 with its rich identity and payment transaction data, through to the complete records of the Government Archive. This matters because language models are bad at business identity. They'll tell you confidently that a business is at an address it left two years ago, or mix up two companies with similar names. When they can query current, verified data instead of guessing from training data, they stop guessing and start knowing. The improvement isn't marginal.
And of course, the Government Archive is now live. It's the government data corpus we've been assembling for years: far beyond corporate registrations; it's the professional licenses, enforcement actions, permits, contracts, shipping records, registered plans, and more that businesses generate as they operate and governments record as they work with these entities. It's all connected to the right business, so it's just a matter of querying for what you need.
The first thing coming in 2026: On-Demand Attributes. You tell us what you need to know about a business. We find it in the government records. Custom structured research, from authoritative and trusted data, delivered programmatically.
Twelve years ago, we showed what was possible when public data gets connected. This year, we rebuilt Enigma to bring those capabilities to AI-backed systems and experts augmented with AI tools. We’re going to keep doing it, because creating trusted business data is who we are. Thanks for building with us.