Ellis AI Raises $10M to Automate Fragmented Private Credit Operations

Ellis AI, a startup building software for private credit managers, announced its emergence from stealth mode Thursday with $10 million in seed funding. The round was led by investors including First Round Capital, 645 Ventures, Harlem Capital, Khosla Ventures, Thrive Capital, Slow Capital, Kearny Jackson, and Ariel Alternatives CEO Mellody Hobson.

The company aims to solve a persistent infrastructure problem in private markets: the fragmented tooling that leaves firms juggling disconnected spreadsheets, document repositories, accounting systems, and email threads. Ellis uses AI agents to integrate these scattered workflows into a single platform, automating routine operations like month-end close procedures and portfolio reporting.

The Founder and His Track Record

Ellis was founded by Ryan Williams, a repeat founder best known for co-creating Cadre alongside Josh and Jared Kushner in 2014. That real estate investment platform became one of the fintech era’s notable exits: it raised more than $160 million across its funding rounds, reached a peak valuation of $800 million, and was acquired by alternative investment company Yieldstreet in 2024 for an undisclosed price.

Williams’ track record provides both credibility and insight into the specific pain points he is now addressing. Having built infrastructure for one segment of private markets, he spotted the operational inefficiencies plaguing another.

“At Cadre, I saw the next major constraint,” Williams said in a statement. “Even as the front end of private markets became more modern and accessible, the operating infrastructure underneath it remained fragmented.” He began developing Ellis last year.

The Problem: Chaos at Scale

Read also: Netflix Invests $587 Million in Ben Affleck’s AI-Powered Filmmaking Platform · Neil Rimer Predicts AI Investment Boom Is Just Beginning · Applications Close in 48 Hours: Your Complete Guide to Stripe’s Startup Battlefield for Australian Founders · The Price of Ambition: How Young AI Founders Handle Relentless Pressure

Manual processes dominate private credit operations

Private credit funds operate with a patchwork of legacy and modern software. Portfolio managers, analysts, and operations teams pull data from multiple systems—custodian platforms, fund accounting software, investor reporting tools, and spreadsheets—then manually consolidate and reconcile the information. The process is slow, error-prone, and resource-intensive.

Excel as a de facto operating system

Williams emphasized this particular dysfunction: “A team may have to download files from several systems, reformat the data, compare balances, investigate discrepancies, and re-enter information by hand. In many firms, Excel becomes the operating system.”

This reliance on spreadsheets creates cascading risks. Version control becomes unclear. Updates in one system don’t propagate. Discrepancies between spreadsheet versions and source systems go undetected until reconciliation deadlines approach. Month-end closes, typically the most data-intensive operational event for a fund, can stretch over weeks of manual work.

The switching cost trap

Traditional vendor solutions demand firms rip out their existing tools and migrate entirely, a costly and disruptive exercise many firms avoid. Ellis takes a different approach: it connects to systems firms already use rather than requiring wholesale replacement.

Ellis’s Architecture and Capabilities

Ellis centralizes data from scattered sources into one accessible interface. The platform uses AI agents to perform specific operational tasks, with a particular focus on month-end fund close operations. The agents can extract files from multiple systems, standardize formats, compare balances across sources, flag discrepancies for human review, and generate reports.

The system also performs ongoing portfolio monitoring, automatically identifying anomalies that might warrant attention. Beyond reactive troubleshooting, Ellis can help prepare investor reports and other deliverables that typically consume hours of manual work.

Williams described the problem Ellis solves as connecting “all the scattered software, accounting information, and documents a private credit firm would use into one easily accessible platform.”

Human Judgment Remains Central

The limits of autonomous AI

Ellis is deliberately designed to keep humans in the decision loop. Williams was explicit about this constraint: “Material decisions and actions remain with the human experts.” The platform is built to augment, not replace, the judgment of experienced operators.

The narrowing human loop

When asked whether he envisioned a future where AI agents operate fully autonomously, Williams offered nuance: “I expect the human loop to become narrower, but not disappear. Our goal is not to replace human judgment; it’s to help people cut through the noise and make educated decisions faster.”

This philosophy reflects a pragmatic middle ground. The agents handle data aggregation, formatting, and flagging, eliminating grunt work. Humans focus on interpretation, judgment calls, and material approvals. Over time, as the system learns and becomes more reliable, fewer decisions may require human intervention—but the ultimate authority remains with people.

Market Timing and Investor Confidence

The funding round demonstrates investor confidence in both the founder and the problem space. The investor group spans established venture firms (First Round, Khosla), emerging funds (645 Ventures, Harlem Capital), specialist fintech investors (Slow Capital), and alternative asset operators (Mellody Hobson of Ariel Alternatives), suggesting broad recognition of the issue.

The private credit market has expanded significantly over the past decade as banks retreated from lending and institutional capital sought direct credit exposure. The $1.5 trillion private credit industry has grown faster than the infrastructure supporting it, creating exactly the kind of operational friction Ellis targets.

Ellis launches into this environment with a specific solution to a concrete problem faced by every growing private credit manager. The $10 million seed will fund product development, hiring, and initial customer acquisition.