Credible Data has raised $10 million in seed funding to expand a platform that provides AI agents and applications with governed definitions and business context for structured enterprise data. The round included Gradient, SignalFire, and K5 Global. Angel investors included G2 co-founder and CEO Godard Abel, pandas creator and Composed Ventures General Partner Wes McKinney, Snowplow Analytics co-founder and CEO Alex Dean, SV Angel and others.
Credible Data is built on Malloy, an open-source semantic modeling language. The platform turns organization-specific metrics, definitions, entities and relationships into executable context that can be queried by AI agents, analytics applications and application programming interfaces.
Enterprise information stored in data warehouses such as Google BigQuery, Snowflake and PostgreSQL is organized into tables, rows and columns. However, the data does not necessarily explain how a company defines concepts such as revenue, an active customer, churn or profit.
Different teams may use different definitions or apply separate business rules to the same information. An AI system with database access but no common semantic layer can therefore produce conflicting answers even when it is querying accurate underlying records.
Credible’s platform analyzes how an organization uses its information, including existing queries and edge cases, to create a governable semantic model. The context can then be supplied at runtime whenever an agent or application needs to interpret company data.
The enterprise product adds security, role-based permissions, governance, reliability controls, auditability and other safeguards to Malloy’s semantic modeling capabilities. This is intended to ensure that systems receive consistent definitions while respecting restrictions on who can view particular data.
Founder and CEO Kyle Nesbit previously led business intelligence and data analytics at Google Cloud and helped integrate Looker into the platform. Head of Product James Swirhun spent eight years at Google working on AI and machine learning products, including Gemini.
VideoAmp is using Credible Data to provide its AI systems with governed definitions for advertising planning and measurement.
KEY QUOTES:
“Every organization attaches its own meaning to its data, even a term as familiar as ‘revenue’ can mean completely different things from one company to the next. AI can’t be trusted with enterprise data until it understands that meaning.”
“The winners in enterprise AI will be the ones who deliver their data’s meaning as context wherever decisions get made, without locking themselves into a single vendor’s stack. That’s what we’ve built at Credible Data.”
Kyle Nesbit, Founder and CEO of Credible Data
“Existing data warehouse and pipeline solutions were never built for AI. As enterprises rush to deploy AI throughout their organizations, they are finding their ability to trust this new technology undermined by conflicting outputs stemming from a lack of consistent meaning applied to their data.”
“Credible Data has built a platform that solves this challenge with unprecedented agility and flexibility, and which we believe will be fundamental in finally achieving trustworthy AI for enterprises.”
Zach Bratun-Glennon, General Partner at Gradient
“Our planning and measurement drives decisions by some of the largest companies in the advertising industry, so every insight our AI gives has to be consistent, defensible and grounded in fact.”
“Credible gives our AI systems the same governed understanding of our metrics that our best analysts carry in their heads. That’s what moved us from experimenting with AI to putting it in front of customers with confidence.”
Peter Nummerdor, Senior Vice President of Product at VideoAmp
