Israeli Startup Catch Raises $5 Million for AI Executive Assistant Agent

Tel Aviv-based startup Catch announced on September 2, 2026 the closing of a $5 million seed round. The company is building an autonomous AI assistant designed to take over managers' and executives' day-to-day administration, such as scheduling meetings, handling email, booking travel, or making calls on their behalf, involving a human only when a decision genuinely requires their judgment.

What Catch Does

Catch integrates with Gmail, Outlook, Google Calendar, Slack, WhatsApp, iMessage, and the phone, without requiring custom workflow setup or any coding. The system observes a manager's activity, learns their preferences, meeting patterns, and travel habits, and then independently carries out routine administrative tasks.

An example given by the company illustrates how this works in practice: when Catch detects a flight booking without accompanying lodging, it checks available hotels in the user's preferred chains on its own and asks for approval of the booking in chat before finalizing it. All financial decisions require explicit user consent, which is meant to limit the risks associated with unsupervised AI agents.

Splinter Agents Architecture

Rather than a single large model with access to all data, Catch relies on what it calls splinter agents, narrowly specialized agents with limited access to information, each assigned to a specific task. The company says this approach improves response quality, shortens response times, lowers computing costs, and reduces security risk, since no single agent ever sees the full scope of a user's data.

The platform holds SOC 2 Type II certification, has passed a CASA Tier-2 audit and Google verification, and encrypts data in transit. Catch also anonymizes user data used for post-training the open-source models it partly relies on.

A Small Round by Design

The $5 million round was reportedly kept deliberately small even though investors wanted to put in more. Most of the funds are earmarked for developing travel features and acquiring users, primarily in the United States, with Europe as a secondary focus. Before writing a single line of code, the founders spoke with more than a hundred executive assistants to understand what their work actually looks like.

Who Are the Customers

Among Catch's current paying customers, the company lists AI startup founders and senior lawyers, people with packed calendars and a high volume of repetitive administrative tasks. CEO Nir Sabato previously invested in early-stage companies at Entrée Capital, where he backed Weavy, a company later acquired by Figma for more than $200 million. CTO Yoav Ramon has more than a decade of experience in machine learning, including speech and language technologies.

The market is flooded with AI agents built for technical people. What sets Catch apart is that the product was built for people who actually run businesses. - Avi Eyal, co-founder of Entrée Capital

Placement in the AI Agent Market

Catch is part of a growing wave of startups building executional rather than merely advisory AI agents, ones that take action themselves instead of just suggesting next steps to the user. That sets it apart from classic productivity assistants, which generate suggestions but don't carry out tasks on the user's behalf without additional integration.

For Polish readers, Catch's story is another signal that the segment of administrative agents for executives is still taking shape and is not yet dominated by big tech companies. Small, focused teams with a specific niche can still raise funding from established funds if they can show real traction measured in thousands of completed tasks, not just announcements.