Why Seda exists

When we started Secoda (searchable company data), we had a clear vision: make company data as easy to search and understand as using Google. Companies were and are drowning in data across dozens of tools, and their teams spent countless hours just trying to find and make sense of it all.

Over the past three years, we’ve built something very powerful in the process. Secoda helps hundreds of companies organize, understand, and search millions of datasets. The "company" part of Secoda was also about solving the complex challenges of company-wide data discovery and access: permissions, governance, documentation, lineage, and more.

But along the way, we’ve built something interesting. Because search has always been our focus, we’ve been able to build an extremely magical AI search that let people talk to their company data naturally.

This meant someone can just ask "What were our top-performing products last quarter?" and get an immediate answer. You could also ask to code, analyze, document or query whatever you wanted. Essentially making anyone capable of working with data.

We've done this by building the infrastructure behind great data search including our own embeddings models, lineage parsing, integrations extractions engine and RAG system that all works around data permissions. With some recent updates, we've had some truly "holy shit" moments with our customers using Secoda AI when working with data. That's why we're thrilled to introduce Seda (searchable data) to the world.

Seda leverages the powerful Secoda backend and the work we've put into this problem over the last three years. This means Seda is complete with advanced lineage parsing, extraction, ranking, and search capabilities to deliver the most accurate responses to your data inquiries. The difference between the tools is that Seda doesn't expose any of those complexities on the front end. By utilizing custom embeddings models, a robust integrations extractions engine, and a sophisticated Retrieval-Augmented Generation system that respects data permissions, the AI can instantly understand and retrieve the precise data needed when you ask, “What were our top-performing products last quarter?”

By dropping the "company" requirement, we're making our AI accessible to anyone with a data source to connect. We believe that the magic of AI search shouldn't be gated behind enterprise deployments and company-wide rollouts. With Seda, you don't need your entire company to adopt a new data platform. You don't need to coordinate across teams or get buy-in from leadership. You just connect the tools relevant to you and start having conversations. The same AI that powers our enterprise product is now available to anyone, without the organizational overhead.

This stripped-down version of Secoda is our reflection of our broader vision for the future of data intelligence and access. We believe that talking to data should be as natural as talking to a knowledgeable colleague. For Secoda customers, nothing really changes. Secoda still continue serving organizations that need comprehensive data discovery platforms, while Seda just opens up the magical core of our technology to everyone else.

Whether you're part of a large enterprise using Secoda or an individual using Seda, we believe you should be able to work better with your tool and your data with our AI. That's what we’re trying to build here and we hope you all find it as useful as we have!