The vast majority of humanity is still not truly benefiting from AI.
That sounds odd in a year when AI assistants are all the rage. Meta’s Muse will book your travel and make purchases on your behalf. Instinct has gone viral in Silicon Valley. But step outside the tech bubble, and we’d argue most people still don’t have a good use for a personal AI agent.
Where these agents can drive real impact is at work. Think about your doctor. You probably don’t want them running your medical information through a consumer AI assistant. But you definitely want them to be AI-enabled, with the admin work handled, so they can give you 100% of their time and focus.
Closing that gap takes a product that feels consumer-grade but also has the security, governance, integrations, and enterprise readiness businesses need. Until Zenzap, we hadn’t seen an AI assistant that does both seamlessly.
The simplest way we can describe it is Instinct or Muse for your professional life.
That’s why Lightspeed is leading Zenzap’s $30M Series A alongside our friends at Bessemer Venture Partners.
Built for the Real Economy
Zenzap customers are in food service, hospitality, healthcare, retail, construction, digital agencies, and defense, and range from small teams to organizations with thousands of employees.
For all the noise, AI has barely diffused into these areas of the economy. Andrej Karpathy put it starkly: around six billion people have barely touched an LLM; another billion or two only use a free agent slightly better than a search engine. For people whose work happens in conversation with others, outside of a terminal, AI hasn’t changed much of their day-to-day. We think part of the reason is that the software they use was never designed for them.
Most of their businesses never adopted enterprise collaboration tools. Slack transformed how Silicon Valley communicates, but it was built around desks and keyboards, and in our view it never reached most businesses outside tech. Instead, these businesses typically use group chats, usually WhatsApp or iMessage.
Colleagues trade personal phone numbers, and the business has no record of or control over the conversations its operations depend on. It isn’t compliant, it isn’t secure, and it produces real problems: consider a nurse photographing a chart to ask a colleague a question, and the image syncs to a personal cloud account. For CISOs and CIOs, it’s a standing nightmare. But the consumer apps are so good that nobody has been able to pry people off them.
Now AI agents are arriving, and the stakes are higher. If businesses don’t give employees a safe, sanctioned way to work with agents, they risk employees bringing their own.
Enter Zenzap
Zenzap spent three years in stealth building a messaging platform that feels consumer-grade, with the low latency and scale people expect from WhatsApp, but with a server in the middle. That architecture is what gives a business central control over users, data, and policies. It’s also what makes the AI useful: agents that join group channels can see the whole organization in context, becoming closer to digital employees anyone in the company can put to work. They sit in shared group chats, connect to the tools a business already uses, from QuickBooks and HubSpot to Procore and Deputy, and take action across systems.
Any employee can create an agent in about 90 seconds without writing code. In one workflow Zenzap describes, a clinic’s agent picks up a patient cancellation and posts it to the reception channel, where a receptionist claims it and fills the slot. In another, an agent drafts an invoice or customer reply and a manager approves it with one tap before anything goes out.
We believe messaging is the right wedge into this market. Chat is a universal need; it spreads naturally within a business, and it holds context about how a company actually runs that exists nowhere else. That makes it the natural home for AI agents.
Consumer-Grade, Enterprise-Ready
Zenzap’s control layer is what makes it possible to put agents in front of a doctor, a store manager, or a line cook. Teams create separate agents for separate jobs, each with access to only the tools that job needs. Each agent runs in an isolated cloud computer dedicated to a single conversation, accepts only cryptographically signed instructions from Zenzap’s platform, and holds no administrator rights or infrastructure credentials.
Admins can require approval before an agent acts, restrict it to read-only, or block actions entirely. Zenzap’s Trust Center lists SOC 2 Type 2, ISO/IEC 27001, GDPR, and HIPAA compliance.
Early Signal
Zenzap reports that thousands of businesses use the product daily across the US, UK, and Australia. We’re also seeing organic pull into some of the world’s biggest chains and enterprises.
The AI agents are following the same pattern. When Zenzap rolled them out to existing customers, the company says 60% of those who turned them on were using them daily within two weeks, having set them up and connected their own tools without help. We think that’s what sets Zenzap apart: a product that can grow like a consumer app and monetize like a B2B company.
Why Guy, Jonatan, and Noam
Zenzap’s founders bring a rare mix of deep security engineering, messaging expertise, and consumer-scale growth. Guy Weiss, co-founder and CEO, led a team in an IDF offensive cyber unit before founding and selling Stucco Media. Jonatan Bien, co-founder and CTO, was Director of Product at Intucell, which Cisco acquired, and VP of Engineering at Voca AI, which Snap acquired, where he worked on Snap’s messenger. Noam Neumann, co-founder and COO, served in an IDF commando unit and later led mobile at Matomy Media.
We believe Zenzap can give every employee in the world, not just engineers at tech companies, the ability to create AI-powered digital workers quickly, safely, and with control. The first wave of AI was personal. We think the next one belongs to the clinics, kitchens, job sites, and stores that run the real economy.
We’re excited to work with Guy, Jonatan, and Noam as they bring AI to the businesses the last generation of work software left behind.
The content here does not constitute an offer to sell or a solicitation of an offer to buy any securities or investment advisory services.
The views expressed are those of the authors and do not necessarily represent the views or opinions of Lightspeed. Other market participants could take different views. Unless otherwise indicated, the inclusion of any third-party firm and/or company names, brands and/or logos are for representational purposes and does not imply any affiliation with these firms or companies and also does not imply their endorsement of the views expressed by the authors.
Certain information contained herein is based on information from various sources prepared by third parties. While such sources are believed by Lightspeed to be reliable, neither Lightspeed nor its affiliates assume any responsibility for the accuracy or completeness of such information, and such information has not been independently verified by Lightspeed. For more details please see https://www.lsvp.com/legal.
Authors
