Modus, an Israeli startup founded in 2026, has raised $10 million in a Seed round led by Insight Partners, with participation from Soma Capital and a group of entrepreneurs and executives — including Eyal Kishon, Nadav Abrahami (Wix, Dazl), and the founders of Cyera and Epsagon. The company is building an infrastructure layer it calls the “Context Warehouse,” designed to improve relevance, latency, and cost for enterprise AI agents by addressing what Modus calls the “Context Gap.”
WHY THE “CONTEXT GAP” HINDERS AI AGENTS
Enterprises today connect AI models to a growing array of sources — tables, documents, code, dashboards and internal conversations — but often still get unreliable or noisy results. Modus argues the issue isn’t a lack of information but too much of the wrong information: agents pull large volumes of irrelevant data and repeatedly query backend systems, driving up compute costs, increasing latency, and complicating scaling.
MODUS: A “CONTEXT WAREHOUSE” TO FILTER THE NOISE
Modus’ platform aims to build and continuously maintain an operational model of an organization’s context. Instead of relying on static documentation or manual definitions, it analyzes recurring analyst queries, the dashboards and reports in use, plus documents, pipelines, code and decision-making conversations. That living representation is updated as the organization evolves and supplies AI agents only the context relevant to each task. Modus claims this approach can reduce token usage and unnecessary data accesses by as much as tenfold.
MODEL INDEPENDENCE AND BUILT-IN GOVERNANCE
The company positions the Context Warehouse as model-, data-platform- and application-agnostic, easing adoption of new models without reengineering how organizational context is managed. The platform supports protocols such as the Model Context Protocol (MCP) and enforces governance rules before information reaches a model, ensuring agents only access data they are authorized to use.
FOUNDERS AND POSITIONING
Modus was founded by CEO Daniel Shimoni and CTO Tomer Mesika. Shimoni previously served as VP of Product at Lusha and brings a conviction that AI effectiveness hinges on business context. Mesika was Head of Architecture at Cyera, working on large-scale classification, governance and protection of sensitive information. The startup currently employs 12 people, all based in Israel.
OUTLOOK
Modus’ funding and technical approach underscore a broader trend: as enterprise data sources multiply, a dedicated context layer could become a critical part of AI infrastructure. The key questions now are how smoothly the Context Warehouse will integrate with existing stacks and whether the promised improvements in relevance, latency and cost hold up in production.
