Illustration for: Brellium Raises $23M To Hunt AI Billing Errors

Brellium Raises $23M To Hunt AI Billing Errors

New York-based Brellium raised $23M led by Left Lane Capital and First Round Capital, with its CortexLM system now auditing more than 22 million patient visits for billing and documentation errors.

By the Numbers

$23M
New round
$16.7M
Apr 2025 Series A
22M+
Patient visits audited
~$2B
Provider revenue 'protected'
ShareXLinkedInEmail

THE RUNDOWN

1

Left Lane Capital and First Round Capital co-leading, with Digital Health Venture Partners joining, shows specialist healthcare investors converging on documentation-compliance AI specifically.

2

The jump from a $16.7M Series A in April 2025 to this $23M round 18 months later is a meaningful but not explosive step-up worth comparing against faster-moving healthcare AI peers.

3

Headspace's reported jump from 2% to 100% note-review coverage is a concrete before/after metric diligence teams can actually verify by asking the customer directly.

4

Brellium's self-reported $2B 'protected revenue' figure is an estimate, not an audited number — worth asking for methodology before citing it as fact.

The VC Read

Value Add VC analysis

The number to press on in diligence is the false-positive rate, not the $2 billion headline: a system auditing 100% of notes is only useful if it doesn't bury providers in false flags, and Brellium hasn't disclosed that figure. Headspace's 2%-to-100% review-rate jump is the most concrete, verifiable data point in this announcement — worth a reference call before taking the billion-dollar protection claim at face value.

Analysis

Brellium, a New York-based healthcare AI startup, has raised new venture funding led by Left Lane Capital and First Round Capital, with participation from Digital Health Venture Partners, according to Tech Startups.

Founded in 2021 by CEO Zach Rosen, Brellium built CortexLM, a system that reviews every patient record and clinical note — not a sample — for documentation errors, billing code mistakes, and compliance gaps before claims reach insurers. Brellium's funding history:

  • Series A, April 2025: $16.7M
  • This round, October 2026: $23M, led by Left Lane Capital and First Round Capital
  • Disclosed funding to date: roughly $40M combined

The Scale Brellium Is Claiming

The company says CortexLM has audited more than 22 million patient visits and helped protect nearly $2 billion in provider revenue from repayment demands, across more than 215,000 providers in 200 provider groups in all 50 states. Named customers include Gentiva, Maxim Healthcare, Rula, and Headspace; Brellium says Headspace raised its note-review rate from about 2% to 100% of visits and now reports 99.2% documentation compliance.

Brellium operates in a documentation-and-coding-compliance niche adjacent to larger healthcare AI players like Nuance, now owned by Microsoft, and Abridge, both of which focus more on clinical note generation than compliance auditing specifically — a distinction that matters for how Brellium pitches itself against better-capitalized rivals.

However, the $2 billion in "protected" revenue is Brellium's own estimate of claims that might otherwise have faced repayment demands, not an audited figure, and the company hasn't disclosed its own revenue or growth rate. Reviewing 100% of notes instead of a sample is a meaningful technical claim, but it says nothing about how often CortexLM flags compliant documentation as an error, which matters as much to a provider's workflow as the headline coverage number.

If Brellium discloses a false-positive rate or lands an independent audit of its $2 billion claim, that would do more for its credibility than any new funding announcement — until then, Headspace's 2%-to-100% review-rate jump remains the most verifiable proof point among the four customers it has named.

ShareXLinkedInEmail

Key Sources

2 sources

THE WIRE in your inbox— Tech, startup & VC news with The VC Read, a few times a week. Free to subscribe, no spam.

No spam, unsubscribe anytime. Sent via Substack · Privacy · or subscribe on Substack ↗