Clearwater Analytics used its Clearwater Connect 2026 conference in Boise to put a number on its AI push. On September 15, 2026, the company said nearly 900 clients now run agentic workflows in production across investment operations, risk and private markets. The Clearwater agentic AI milestone arrived alongside a research report, GenAI and the Data Divide, which surveys institutional investors on why AI pays off for some firms and stalls for others.

The count is real, and it is also broad. Read closely, the company release gives enough detail to judge how deep the usage goes.

What the Clearwater Agentic AI Numbers Show

Clearwater describes the nearly 900 as clients using its AI capability. That covers conversational AI, automated workflows and specialized agents. Within that group, the number of clients building their own custom workflows rose 197% since January. The number running workflows rose 69% over the same period.

Clearwater agentic AI workflow runs are up 105% since January. In one recent week, clients completed 3,591 runs with a 96% success rate. Those are production numbers, not a pilot cohort.

The agent count has also moved on. Clearwater put nearly 1,000 AI agents on its platform when it reported fourth-quarter 2025 results in February, and its website now says more than 1,000. Earlier figures of around 800 are out of date.

For scale, Clearwater lists about 2,400 clients on its website. On that basis, more than a third of the client base now touches Clearwater agentic AI in some form.

Built Into the Record, Not Bolted On

Clearwater sells investment accounting, reporting and risk software to insurers, asset managers, banks, corporations and governments, and says it supports more than $10 trillion in assets. The back-office work it handles is reconciliation, performance and the reports boards and regulators expect.

Every Clearwater agentic AI capability runs against the same continuously reconciled investment record that powers that accounting. Outputs trace back to source data, with citations in conversational answers and step-level detail in automated workflows. In February, Clearwater also said it had embedded agentic AI in its Beacon risk platform, working on live portfolio data inside the core calculation engine.

Two clients gave concrete examples of Clearwater agentic AI at work. Blue Cross Blue Shield of Michigan uses CWIC Flow to automate the first review of a monthly validation that spans 198 accounts across six entities. Securian Asset Management says it uses CWIC to cut manual work and automate workflows. Clean, structured data is a real advantage here over vendors trying to automate messier environments.

The Survey Supports the Pitch

The research report does not complicate Clearwater’s story. It is built to support it. The company frames the survey and the Clearwater agentic AI milestone as opposite sides of the same divide: firms struggle because their data is unreliable, and its clients succeed because theirs is not.

The headline numbers are striking. Ninety-five percent of institutional investors call AI important to their goals, and 79% expect AI-enabled firms to outperform. Yet only 40% say AI informs more than a quarter of their operational decisions. Some 79% rate their data as complete, but only 56% rate it as accurate. And 99% name unreliable or opaque data as the biggest barrier to trusting AI outputs.

CEO Sandeep Sahai made the link explicit: “Only 56% of firms trust the accuracy of their own data.” The findings are useful. Still, this is a vendor survey whose conclusion matches the vendor’s product, and the release does not give the sample size. Readers should weigh it that way.

Private Now, With Less to See

Clearwater is no longer a listed company. Its take-private by an investor group led by Permira and Warburg Pincus closed on June 25, 2026. The deal was valued at about $8.4 billion, or $24.55 per share. Francisco Partners supported it, and Temasek took part. The stock left the New York Stock Exchange the same day.

That changes how milestones like this one reach the market. There are no more quarterly earnings calls where analysts can press on adoption figures. The Clearwater agentic AI numbers now arrive through press releases and conference stages, which the company controls.

Going private was pitched partly as a way to fund Clearwater agentic AI development. Sahai said at closing that private ownership would help the company build its agentic platform while scaling the current one. The September milestone is the first large public test of that promise. It is also unaudited, which is normal for a product release but worth keeping in mind.

What Would Show Depth, Not Breadth

The obvious question is whether nearly 900 clients means deep use or generous counting. Clearwater has already published some of the breakdown needed to answer it.

A simple division helps. Clearwater reported 3,591 workflow runs in a recent week across nearly 900 clients, which averages about four runs per client per week. Usage will vary widely, but an average that low suggests most Clearwater agentic AI use is still light, targeted automation rather than firm-wide transformation.

Narrow automation can still be useful, as the Blue Cross example shows. The next disclosure worth watching is the distribution: how many clients run workflows every week, and which tasks account for most runs. A named client describing agents in risk or private markets work, beyond validation reviews, would show whether Clearwater agentic AI is moving past the back office.

For related coverage, see how Nasdaq Ventures backed an AI governance layer for capital markets firms, why Banyan Software bought into the WIZE wealth platform, and what MSCI bought with First Street climate data.

FintechBits covers payments, banking and fintech developments for readers in the US and UK. This article is for information only and is not financial advice. Views expressed are those of the FintechBits editorial team.