About CleanAim®

33 years of learning systems. 8 patents. 1.1 million lines of proof.

CleanAim® builds infrastructure that makes AI systems provably compliant, continuously improving, and portable across providers. We're solving the same problem we've been solving since 1993: how do you capture decisions, pair them with outcomes, and enable compound improvement?

Systems that learn from outcomes always outperform systems that optimize once

CleanAim's architecture originated from a simple insight that has driven four companies over three decades.

1996–2004
x.hlp Technologies Human learning systems, acquired by SAP
2005–2013
Trading Systems Financial prediction infrastructure
2014–2025
Wyetree Energy coordination software
2020–present
CleanAim® AI governance infrastructure

The architecture evolved, but the core principle remained: capture decisions, pair with outcomes, enable compound improvement.

Optimization and compliance are the same problem

When large language models emerged, we recognized the opportunity: the same architecture that enables continuous improvement also generates the audit trail that compliance requires.

A system that learns from outcomes naturally produces the evidence regulators need.

Most governance only observes AI. CleanAim® learns from it.

We didn't just design it. We built it.

1.1 million lines of production code, all governed by the same framework we offer to customers. The system is honest enough to say "insufficient evidence" rather than certify what it can't verify.

1.1M

Lines of production code

98/100

Audit score—2 checks await calibration data

411

Learned patterns, 389 auto-promoted

0

Blockers—zero architectural violations

Tom Killi

Founder & CEO

Tom Killi has built systems that learn from the prediction-outcome loop since 1993.

In 1996, he founded x.hlp Technologies, which developed capture-pair-learn methodology for enterprise training. The company served 30+ Fortune 500 clients—JP Morgan, Shell, McKinsey, NATO—before acquisition in 2004. That technology now operates inside SAP.

His prior patent work has been cited by 52+ patents from Microsoft, Apple, Samsung, IBM, and SAP over 25 years.

1.1M+

Lines of Code

52+

Patent Citations

33

Years in Learning Systems

Making AI trustworthy

To build infrastructure that makes AI trustworthy—not by limiting what AI can do, but by ensuring that what it does can be understood, verified, and improved.

Learning Over Documentation
Documentation describes what should happen. Learning captures what actually happens.
Infrastructure Over Policy
Policies can be ignored. Infrastructure cannot be bypassed.
Independence Over Integration
Provider tools assess provider systems. Independent tools assess independently.
Honesty Over Perfection
A system that admits uncertainty is more trustworthy than one that claims certainty.

How we build

Four principles guide everything we create.

Constitutional

We build systems that enforce their own rules. Not guidelines that can be ignored, but architectural constraints that cannot be bypassed.

Honest

Our systems admit when they don't know. A 98/100 score with explanation is more trustworthy than a 100/100 score that papers over gaps.

Independent

Provider tools assess provider systems. We build tools that assess any system from any provider—true independence by architecture.

Compound

One-time optimization is valuable. Continuous improvement is transformational. We build systems that get better every day.

Company facts

Legal Entity CleanAim® Holdings Inc.
Location New York and Munich
Founded 2020
Platform Version 2.5.0 (GA Release)
Patents Filed 8

Get in touch

Whether you're interested in the CleanAim® Platform, partnership opportunities, or just want to learn more.

hello@cleanaim.com