Code suggestion
Diagnosis and procedure codes drawn strictly from the signed documentation, with the supporting text highlighted for the reviewer.
07 · Clinical documentation
AI Coding & Charge Agent
Coding is where accuracy and revenue collide. The agent does the mechanical work at machine speed, and a certified coder owns every judgement call before the claim leaves the building.
Agent at a glance
- StatusIn development
- CategoryClinical documentation
- Deployment2–4 weeks, shadow mode first
- Review modelHuman approval per action, configurable
- Headline metric95%+ Accuracy
- EvidenceFull action log, exportable any time
95%+
Coding accuracy
Measured against certified-coder review on a 10% monthly sample.
< 24 hrs
Note to coded claim
No more charge lag accumulating over a weekend.
< 5%
Coder override rate
A falling override rate is how we track the agent getting better.
What it does
Everything in this agent, in writing.
Modifier logic
Global periods, bilateral, multiple procedure, assistant surgeon and distinct-procedure rules applied with an explanation for each.
Payer rule checks
National and payer-specific edits, medical necessity policies and frequency limits applied before the claim is built.
Charge capture reconciliation
Scheduled visits, procedures and no-charge encounters reconciled so nothing goes unbilled.
Documentation queries
Where the note is ambiguous, a specific query is raised to the provider instead of a guess being coded.
Human coder review
A certified coder reviews, corrects and releases — with every override logged as training signal.
Deployment
Live in weeks, not quarters.
01
Baseline accuracy test
We code 100 of your recent encounters blind and compare to what was billed, so you see the delta before you commit.
02
Rule and fee schedule load
Payer edits, fee schedule and specialty policies configured.
03
Shadow coding
The agent codes in parallel with your team or ours; differences are adjudicated and documented.
04
Production with audit
Straight-through coding for routine encounters, human review for the rest, monthly accuracy reporting.
Guardrails
What it will never do.
These limits hold regardless of configuration. They exist so that you can put an agent in front of patients without putting your licence next to it.
- Codes only what the signed note supports — it cannot invent documentation to justify a code
- Every suggestion cites the note text that supports it, for the reviewer and for an auditor
- Below-threshold confidence always routes to a human coder rather than releasing
- Monthly accuracy sampling by our audit team, reported to you with the misses shown
- Model and rule-set versions recorded per encounter so a historical decision can be reconstructed
Human in the loop, by your choice
Every action can require approval before it commits. Start with all of coding agent queued to your staff or to ours, watch the accuracy numbers, then widen autonomy workflow by workflow — or leave it queued forever. Both are supported.
Escalations land with certified people
Our review queue is staffed by the same certified coders, billers and credentialing specialists who run our managed services. Not a support desk, and not a chatbot watching a chatbot.
Full evidence trail
Input, decision, action, confidence, model version, reviewer and timestamp — retained and exportable as CSV or JSON, including after you leave us.
What we report against
Operational targets for this agent. They are internal commitments measured on your own data, reported monthly — including the misses.
95%+
Accuracy
< 24 hrs
Turnaround
< 5%
Override rate
< 2%
Denial rate
Works with your systems
Your EMR, your numbers, your data.
- Any EMR with a note feed
- Clearinghouse APIs
- Encoder and code-checker tooling
- Athena
- eClinicalWorks
- NextGen
We integrate through FHIR and HL7 where the vendor supports it, and through secure automation or structured document intake where it does not. No API required, no vendor commission taken.
FAQ
Common questions.
Is an AI allowed to assign codes?
The agent suggests; a certified coder decides. That is both better governance and better economics — and it is how we would defend the record in an audit.
How do you handle risk adjustment and HCC?
Chronic conditions are captured with MEAT documentation review, hierarchy management and a query process where documentation supports a higher-value diagnosis but does not state it clearly. Suspect conditions are never coded without provider confirmation.
What happens when it disagrees with our coders?
Every disagreement is adjudicated and recorded. If the agent is right, we fix the rule. If your coder is right, the agent learns. You get the override rate monthly.
Does this replace our coding team?
It replaces the mechanical part of their job — code lookup, edit checking, charge reconciliation — and leaves them the judgement. Most practices redeploy rather than reduce.
Related agents
Usually deployed together.
08
In development
AI Intake & Eligibility Agent
Collects history, consents and insurance before the visit, verifies eligibility and benefits in real time, and writes clean structured data into the right EMR fields.
- Digital intake completed before the patient arrives
- Eligibility, deductible and co-insurance verified in real time
- Clean structured data written back — no retyping
09
In development
AI Recall & Care Gap Agent
Works your recall list and quality gaps continuously: overdue screenings, missed follow-ups, annual visits and chronic care check-ins — outbound, in your voice, with the appointment booked.
- Overdue recalls and quality gaps worked continuously
- Outbound outreach that ends in a booked visit
- Closes HEDIS and MIPS gaps with the documentation to prove it
10
In development
AI Prior Authorization Agent
Checks whether a service needs authorization, assembles the clinical packet, files it, and tracks every request to a decision — with the deadlines that matter.
- Requirement checked before the service is scheduled
- Clinical packet assembled from the chart automatically
- Every request tracked to a decision, with deadlines enforced
See it run on your own workflow.
A 30-minute demo using your scenarios, your payer mix and your systems. We will show you the guardrails and the audit log, not just the happy path.
A 30-minute call with an operations lead. No obligation, no sales sequence, and we will say so if you do not need us.