Ekai raises $1.7M to give enterprise AI agents verified business context
SiliconANGLE Duncan Riley ● Covered by 3 sources
Ekai raised $1.7M to help AI agents use company data with the right business meaning. It says verified definitions can cut modeling work from months to hours.
Based on reporting by SiliconANGLE, Duncan Riley — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Ekai Inc. has raised $1.7 million to push a very specific idea: AI agents should not be trusted with corporate data until the business meaning behind that data has been checked by humans who actually know the business.
The company’s platform builds the semantic models and transformation code that let AI tools read a warehouse correctly. Ekai starts with domain experts. They define what a term or metric means, and those definitions become the source of truth before the system turns them into machine-readable logic and validation rules. Nothing gets shipped until it has been checked against the warehouse data, the company says.
That is the part Ekai calls forward-engineering. It is a jab at a lot of the market, which the company says works in reverse by trying to infer meaning from old BI dashboards, query history and dbt projects. Co-founder and CEO Moatassim “Mo” Aidrus put it bluntly: reverse-engineering is like asking the exhaust pipe what the engine was thinking. Useful for documentation. Not much help for what comes next.
Co-founder and Chief AI Officer Hussnain Ahmed argues the real problem begins with the models. A foundation model may know the phrase “active user,” but it does not know what that means inside one company or where that meaning lives in the warehouse. Ekai wants each definition captured from the people who own it, and signed off by name.
The company says that approach has sped up early work dramatically. Semantic modeling that often takes teams three to six months was done in as little as six hours in some engagements, with verification doing the heavy lifting. And while the AI world keeps talking about memory, retrieval and prompts, Ekai is staying a layer below that: first verify the business meaning, then let the agent reason over it.
The pre-seed round was led by Misneach, with Cambridge AI venture fund and C10 Labs also participating. Ekai says the money will go toward product development, go-to-market work and deeper integrations. Its workflows are already on Snowflake, including the Snowflake Marketplace, and also support Databricks, BigQuery, Redshift, Azure Synapse, Postgres, ClickHouse and DuckDB.
My take — AI-written commentary, not fact-checked reporting
This is the unsexy part of AI that keeps getting skipped, which is exactly why it matters. Everyone wants agents; fewer people want to admit their data still needs a grown-up in the room. Ekai is betting that verified business meaning will become more valuable than fancier prompting, and that sounds less like hype than basic hygiene.
Read more about this at: SiliconANGLE