Cierta AI
Evidence-grade risk-adjustment coding

Chain of custody
for medical coding

Cierta ties each supported code to source documentation, routes it through human review, and seals the trail in a tamper-evident record. Defensible only when the evidence, approval, and chain of custody are present.

Request a demo See how it works
Synthetic-first Human-in-the-loop Tamper-evident seal

Forensics has chain of custody. Finance has the audit trail.
Risk-adjustment coding has had neither — until Cierta.

The problem

A code you can't defend is a liability

Risk-adjustment and HCC coding face growing audit scrutiny. When a finding can't be traced to the record, it becomes exposure — not value. Most tools return a score; few can show their work.

Audit pressure is rising

RADV and payment-integrity reviews increasingly demand documentation support for every adjustment. "The model said so" is not a defense.

Black boxes don't hold up

A confidence number without an evidence trail can't be reviewed, explained, or stood behind in front of an auditor.

Manual review doesn't scale

Defensible review by hand is slow and inconsistent. Teams are forced to choose between coverage and rigor.

How it works

Evidence first, by design

Cierta turns a chart into reviewable, traceable findings — with a person in the loop at the decisions that matter.

Ingest

Charts enter a controlled pipeline; real-PHI client activation stays gated until agreements and approval are complete.

Analyze

The platform extracts conditions and candidate codes, citing the exact source pages it relied on.

Evidence-link

Each finding is tied to documentation and screened for MEAT-style support before it's surfaced.

Review & trail

Coders confirm or reject; every action is written to a tamper-evident audit trail.

Capabilities

Built to prove what's true

Designed around defensibility — not just a number, but the reasoning and record behind it.

Evidence-linked findings

Every finding references the source page and provider attribution it came from, so reviewers can verify it directly.

MEAT-aware screening

Findings are screened for Monitoring, Evaluation, Assessment, and Treatment support before they reach a coder.

Human-in-the-loop

Cierta supports coders — it doesn't replace them. Submissions stay behind explicit human approval.

Tamper-evident audit

Actions are recorded to a hash-chained log designed to make the review history verifiable after the fact.

Audit-defense framing

The workflow is oriented toward standing behind coding decisions — not toward maximizing capture.

Explainable exports

Outputs carry their supporting evidence and provenance, so findings travel with the reasoning behind them.

The Cierta Seal

One mark for evidence-backed support

When a finding clears its evidence gates and a coder confirms it, Cierta seals the record: tamper-evident, hash-chained, and verifiable later. It separates "the model is confident" from "here is the proof, and it has not changed."

  • Bound to the source documentation it was drawn from
  • Signed off by a named human reviewer
  • Hash-chained, so any later change is detectable
How we're different

Scoring vs. chain of custody

Everyone else returns a score. Cierta returns a sealed evidentiary record.

Black-box scoring

  • Returns a code and a confidence number
  • No traceable link to the record
  • Hard to review, explain, or appeal
  • Optimized for capture volume
  • History is opaque after the fact

Cierta Evidence-first

  • Every finding cites its source documentation
  • MEAT-style support screened before surfacing
  • Reviewable, explainable, evidence-backed
  • Oriented toward audit defense
  • Tamper-evident, verifiable history
Real-PHI gateSynthetic first; real PHI waits for signed agreements
Organization-awareBuilt around role and organization boundaries
Hash-chained auditTamper-evident record of every action
Activation controlsReal-PHI use requires formal approval

Read more on security & compliance →

Who it's for

For the teams who have to be right

Health plans

Strengthen risk-adjustment integrity with findings you can trace and defend in audit.

Provider groups, MSOs & ACOs

Bring rigor and consistency to chart review without sacrificing throughput.

Compliance teams

Move from black-box scores to a reviewable, evidence-backed record of every decision.

FAQ

Questions, answered plainly

What does Cierta AI do?
Cierta is AI-assisted chart review for risk-adjustment and HCC coding. It surfaces candidate findings, links each to the documentation that supports it, screens for MEAT-style support, and records every action to a tamper-evident audit trail — so coding decisions can be reviewed and defended.
Does Cierta replace coders?
No. Cierta is built for human-in-the-loop review. Candidate findings are presented for a coder to confirm or reject, and submission stays behind explicit human approval.
How does Cierta handle PHI?
The public launch posture is synthetic-data-first. Real-PHI client activation requires signed agreements, compliance approval, and operational activation. See Security & Compliance.
Is Cierta available today?
Cierta is deployed for controlled demonstrations and validation under a synthetic-data-first posture. Real-PHI client activation requires signed agreements and compliance approval. Request a demo to see it on representative charts.

See it on your own workflow

We'll walk you through Cierta on representative charts and show the evidence trail behind every finding.

Request a demo