Daniel Shimoni and Tomer Mesika publicly launched Modus on July 29 with a $10 million Seed round and a bet that connecting AI agents to enterprise data solves only half the deployment problem.
CTech reported on July 29 that Insight Partners led the financing, with Soma Capital, Eyal Kishon, Nadav Abrahami of Wix and Dazl, the founders of Cyera, and the founders of Epsagon participating. The valuation, headcount and customer roster were not disclosed.
CTech describes Shimoni and Mesika as former Cyera and Lusha executives. The launch report does not specify which founder came from which company or their exact titles.
That background gives Modus a product and data-security frame for its pitch: enterprise agents need help determining which information matters for a task, rather than simply gaining access to additional systems.
Turning business context into infrastructure
Modus calls its product a "Context Warehouse." CTech describes the underlying problem as a "Context Gap": models can access corporate tables, documents, source code and workplace conversations while still missing the accepted definitions and business logic needed to use that material reliably.
According to CTech, Modus aims to reduce irrelevant retrieval and the tokens and computing resources consumed by enterprise agents.
A finance agent, for example, may have access to several tables containing revenue figures. Access alone does not tell the agent which table finance leadership accepts, whether a regional team applies a different definition, or whether a recent acquisition changed the calculation. Those distinctions often live across code, dashboards, documentation and messages rather than in one authoritative database field.
Modus claims the Context Warehouse can reduce irrelevant retrieval and token consumption by as much as tenfold. That is a company-provided figure, not an independently benchmarked result. It also spans two operational measures: how much irrelevant data an agent retrieves and how many tokens it spends processing that data.
MCP made connection easier, and raised the stakes
Anthropic introduced the Model Context Protocol on Nov. 25, 2024 as an open standard for connecting AI assistants to systems where data resides, including content repositories, business tools and development environments. Anthropic's MCP made it easier for agents to connect to business systems, which increases the need for relevance and governance layers around what those agents retrieve.
That connectivity created the opening Modus is pursuing. Once an agent can query many systems, companies must still determine which information belongs in its working context, which definition governs the task and which records the user is authorized to see.
Investors are funding several versions of that thesis. Israeli-founded Jedify raised a $24 million Series A in June for a continuously updated "context graph" spanning structured and unstructured business data. Atlan's Context Engineering Studio packages business definitions, rules, evaluations and versioned context repositories for AI agents.
Modus is entering an active category. Its Context Warehouse will have to select useful business information, keep definitions current and produce measurable accuracy gains without forcing data teams to maintain another manual catalog.
The founders are selling the control point
The $10 million round gives Shimoni and Mesika capital to develop the product and hire. The investor lineup also carries relevant operating history: CTech names the founders of Cyera and Epsagon among the participants.
Modus is applying an infrastructure approach to the information an agent receives before it reasons or acts. That puts Modus near data governance, security and identity vendors that are extending their products to cover AI agents.
Cyera has moved in that direction from the security side. CTech reported that Cyera acquired AI data-governance startup Ryft in April for an estimated $100 million to $130 million, a deal that pointed data security vendors toward agent data access and governance.
The activity around context, access and identity reflects a practical problem for enterprise agents. Each additional connection creates another decision about relevance, permissions and governance. Modus is betting that companies will buy a dedicated system to make those decisions before a model consumes the data.
The Seed round gives Modus the resources to test that claim. Modus' value will depend on whether the Context Warehouse improves completed work and preserves access controls, rather than merely reducing the number of tokens required to produce an answer.