Jedify, a New York-based startup building what it calls a context graph for enterprise AI agents, raised $24 million in a Series A round led by Norwest. The deal brings the company's total funding to roughly $33 million and arrives as enterprises wrestle with the gap between AI vendor demos and agents that actually work inside a real business. Snowflake joined as a strategic investor and is integrating Jedify with Cortex AI, Semantic Views, and CoWork.
S Capital VC and Cerca Partners returned from prior rounds, and Oceans Ventures came in as a new backer. Jedify says it has between 10 and 20 early customers, with The Weather Company among them, and is targeting mid-market and large enterprises that already run multiple databases or warehouses. The company will spend the new capital on product, hiring, and go-to-market.
The pitch is straightforward. AI agents inside an enterprise need to know how the company defines revenue, which employees can see which file, and what a given workflow actually involves. Jedify connects via APIs to databases, data warehouses, SaaS apps, BI tools, and unstructured sources like documentation, code, Slack channels, and meeting recordings, then assembles a graph that captures relationships across entities, data, people, and permissions.
Key facts
- 01Jedify raised $24M in Series A funding led by Norwest, bringing total funding to about $33M.
- 02Snowflake joined as a strategic investor and is integrating Jedify with Cortex AI, Semantic Views, and CoWork.
- 03The New York-based startup has 10 to 20 early customers, including The Weather Company.
- 04S Capital VC and Cerca Partners returned for the round; Oceans Ventures came in as a new investor.
- 05Jedify's platform builds a multi-dimensional context graph spanning data, permissions, workflows, and company terminology.
Co-founder and CEO Assaf Henkin pointed to compliance company Kiteworks as a representative deployment. Kiteworks plugged Snowflake, Tableau, Notion, and internal playbooks into Jedify, then built agentic tools for sales workflows.
“They wanted to arm their sellers and account teams with a sophisticated app — you can think of it as both like a dashboard application and a real-time conversational application. When they go into a customer conversation, Jedify builds for them, on the fly, everything they need to know.”— Assaf Henkin, Co-founder and CEO of Jedify
Henkin argues the context graph is structurally different from the semantic layers, metadata catalogs, and knowledge graphs that already populate enterprise data stacks. It is multi-dimensional, model-agnostic, and updates in real time as data moves through the connected systems.
That distinction matters most when agents are asked to act rather than just retrieve.
Permissions are the obvious hazard. An agent that hands a CFO's revenue projections to an intern is worse than no agent at all. Jedify inherits permissions from identity systems, file systems, SaaS tools, and databases, including row-, column-, and table-level rules, and lets customers layer additional groups defining what each agent or workflow can reach. The platform also offers observability and governance tooling to monitor agent behavior.
The Snowflake investment is notable because the large data platforms are themselves racing to build native context capabilities. Henkin's argument is that companies rarely keep all of their data, or their institutional knowledge, inside a single cloud provider's environment. He framed that as a structural disadvantage for hyperscalers pitching an all-in-one stack, and a wedge for a neutral layer that sits across providers.
He also pointed to cost. Training a model in-house to build a comparable context layer can be prohibitive at a time when finance teams are scrutinizing AI token consumption. Jedify's bet is that as frontier models become more capable and more interchangeable, the proprietary context that makes them useful inside a specific business becomes the durable moat — not the model itself.
Interest is concentrated in data-heavy sectors including gaming, industrials, and consumer packaged goods, according to Henkin. Those are industries where institutional knowledge tends to live in a sprawl of documents, telemetry, and tribal practice rather than a single warehouse, which is precisely the territory Jedify is trying to map.
The skeptical case is that context graphs are an early category with several adjacent terms — semantic layers, knowledge graphs, metadata catalogs — already in use, and customers may decide their incumbent data vendor's built-in version is good enough. Snowflake's parallel work on Cortex AI and Semantic Views is the clearest example. Jedify's counter is that its product is complementary, not competing, but that argument has to survive contact with procurement teams that prefer fewer vendors.
If the context-layer thesis holds, Jedify is positioning itself at the chokepoint between frontier models and the messy reality of enterprise data. The AI agent market has spent the past 18 months discovering that raw model capability is not the bottleneck — context, permissions, and governance are. A $24M Series A is small relative to the model labs, but it buys a seat at a layer of the stack that every agent vendor will eventually need to solve, and Snowflake's check signals at least one major data platform would rather partner there than build it twice.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
