Why We Raised $30M to Put Enterprises in Control of AI Agents

Today Kindo announced a $30 million raise co-led by Crosspoint Capital and Forgepoint Capital, bringing our total funding to $57.6 million. TFX Capital and CircleRock Capital joined, alongside our existing investors. We'll use it to grow our go-to-market team and accelerate the run anywhere AI agent management platform enterprises are already running in production.

I want to explain why this matters, and why now.

Revenue grew roughly 800% year over year in 2026. That number isn't the story. The story is what's driving it: in AI forward companies the teams responsible for AI have stopped asking whether agents belong in the enterprise. They're asking how to stay in control of thousands of them.

2026 was the year AI started doing the work

I've watched this movie before. As CIO and CISO at Bird, Clover Health, and Riot Games, I saw mobile, SaaS, and cloud each reach the enterprise years before the tools to centrally control them. When we started Kindo in the Summer of 2022, as generative AI launched, I didn't want Security, IT, Compliance and Finance to wait again. 

This wave is moving faster than any of the previous ones. Agent builders from the AI labs, the hyperscalers, and thousands of startups make it easy to put powerful agents to work. They investigate alerts, update customer records, approve requests, and change production systems, around the clock. Many are built by people who have never shipped software let alone software in the enterprise.

The same permissions that let an agent get work done let it expose sensitive data, act outside policy, break a critical workflow, or run up a huge inference bill. Now that agents are running 24x7 and many organizations are now dealing with thousands of them.

What's missing is tooling that lets Security, IT, Compliance, and Finance centrally control how agents are built, what data they reach, where that data goes, and what they cost. Support teams also need a way to reach in and fix or optimize an agent at 3 a.m.. Without these capabilities, the people responsible for supporting and controlling AI are right to put the brakes on AI transformation.

That's the gap enterprise AI agent management fills.

What enterprise AI agent management looks like in practice

Kindo is the enterprise AI Control Center. It gives Security, IT, Compliance, and Finance one place to discover, observe, and control every AI agent running across the business, whether it was built in Kindo or somewhere else.

1. Discover. Connect the agents, data sources, and models you already use. Kindo maps where agents are in use, including the shadow AI nobody approved.
2. ‍Observe. Every agent's activity, access, cost, and outcome lands in one view. Our gateways govern and audit each call an agent makes to data, tools, and approved models.
3. Control. Identity controls apply your existing permissions to people and agents alike, and let you tighten them centrally. You decide what employees can do with an agent, what an agent can touch, and what it can spend.
4. Operate. For mission-critical agents at scale, Kindo's harness adds 24x7 monitoring, support, and direct intervention, including taking over an agent when something goes wrong.

Put together, that's an AI Operations Center: SOC- and NOC-style monitoring and intervention, built for AI. Implementation is turnkey and run by the IT operations staff you already have. No software engineers required.

And because it isn't "your" AI if it depends on one vendor's API, Kindo works with any model, any agent builder, and any data source. It runs on premises, in your cloud, in Kindo’s cloud, or air-gapped.‍

AI agent cost control is now front and center

Security and compliance were the first reasons enterprises needed control of their agents. Cost is now front and center. AI spend keeps climbing year over year, and it has become one of the biggest things slowing AI programs down.

Every AI program eventually has to answer one question: how do I pay for AI judgment, not routine work?

Kindo gives Finance a way in. We record every token consumed and every agent run's outcome, including the failures. Cost and return on AI spend roll up by person, agent, department, and project. Finance can set budgets, trigger alerts, and enforce hard spend caps. They can see which agents fail often, deliver little, or run on a model that's overkill for the job. Finance has the real time context to ensure spend is driving results.

Kindo’s Deep Hat add-ons drive those costs down saving as much as 95% or even eliminating the need to use AI for the task:

• Deep Hat Sift learns in real-time which events on busy alert and ticket queues an agent would dismiss. Once a person approves the threshold, it skips those runs, logs everything it filters, and samples some through to measure misses.
• Deep Hat Tune automatically adapts a smaller, lower-cost model to a specific task and saves the big models for work that needs them.
• Deep Hat Shrink converts and runs repeatable agent steps as code, which takes model cost to zero, and still routes exceptions back to a model.
• Deep Hat Attack gives security teams powerful low-cost, private offensive security testing and validation, so day-to-day security work doesn't burn frontier model spend.‍

Proven in production

Kindo is trusted and runs in production at large and mid-market enterprises in financial services, global professional services, neoclouds, healthcare, defense, and aviation. In May 2026, the CISO of Aireon, the global aviation in-flight data provider, reported that using Kindo for threat hunting and Tier 1 and Tier 2 SOC work helped the team move faster, get more from its existing security monitoring, and save more than $2 million a year.‍

Why Crosspoint and Forgepoint, and what comes next

We chose partners who understand what it takes to secure the enterprise. Andre Fuetsch of Crosspoint Capital and Ernie Bio of Forgepoint Capital are joining our Board. Andre put it simply: 2026 has been the year of the agent, and most companies still lack a reliable way to see what their agents are doing. Ernie described the path every CIO is on, from one agent triaging security tickets to agents everywhere, and the need for one place to manage access, support, and cost.

I'm also grateful to TFX Capital and CircleRock Capital for joining, and to Drive Capital, Riot Ventures, RRE Ventures, Eniac Ventures, Marlinspike Partners, New Era Ventures, and Sunset Venture for backing us again.

Here's where the money goes:

• Go-to-market. More people to help Security, IT, Compliance, and Finance teams get control of their agents fast.
• The AI Control Center. Deeper discovery, observability, and control across every agent, model, and data source.
• Deep Hat. New models and add-ons that keep driving the cost and risk of running agents down.‍

Hand AI the work, not the keys

Charlie, Tony, and I started Kindo so the people who protect the enterprise never have to choose between going faster with AI and staying in control of it. This raise lets us bring that to many more of them.

If your teams are deciding how far to let agents into the business, I'd like to show you what enterprise AI agent management with run anywhere Kindo looks like.

Get a demo of Kindo.

Read the full announcement: Kindo Secures $57.6 Million to Expand Enterprise Control Over AI Agent Actions and Reduce Costs

-Ron