AN Alpesh Nakrani
SolutionsBlogBooksPraiseAbout Work with me ↗
Solutions/AI denial & appeals agent

An AI denial appeals agent your clinicians sign.

Appeals win four times out of five, and almost nobody files them. I build the machine that removes the drafting hour. The judgment stays with your clinician.

30 minutes. The price on this page is the price on the call.

Appeals Desk illustrative
CO-50 · medical necessity awaiting sign-off
Appeal draft · claim #4821 · lumbar MRI
policy L34220 §2.1 chart note 03-14 grounding 96%
#4795 · CO-197 · auth absentdrafted
#4788 · CO-50 · medical necessitydrafted
#4771 · CO-16 · missing infodrafting…
RE: Appeal of denial · claim #4821 citations verified

“Per LCD L34220 §2.1, MRI is indicated when conservative therapy has failed…”

2 policy citations 4 chart references clinician signature required
Denial appeals evaluation dashboard showing 96.2% citation accuracy, 94.8% factual grounding, 128 of 131 regression checks passed, and zero automatic submissions

Golden set: your historical denials · thresholds are contract acceptance criteria · illustrative values

Letter signed · routed to staff for submission
Clinician gate: enforced in code No auto-submit · ever
81.7%
of appealed MA denials fully or partially overturned
KFF analysis of CMS data, 2023
11.7%
of denials ever appealed
KFF analysis of CMS data, 2023
3.2M
Medicare Advantage prior-auth denials in one year
CMS, 2023

Read that from the revenue side: the overwhelming majority of appealed denials succeed, and the overwhelming majority of denials are never appealed. The money is not in fighting harder. It is in appealing at all.

The problem

The denial backlog is a profit center you’re not running

The industry spends $25.7 billion a year on denial rework, at roughly $57 of administrative cost per denied claim. Initial denial rates average 11.65% and pass 20% in specialty settings. Every RCM firm and specialty practice knows this. The response is always one of three things:

By hand

Write appeals manually

Works, wins often, doesn’t scale. The person who writes good medical-necessity arguments has a full-time job already.

doesn’t scale
Rent

Rent an enterprise platform

Waystar, AKASA, Innovaccer: capable systems, with hospital-system contracts and hospital-system sales cycles. None of them publishes a price.

no public price
Outsource

Hand it to an appeals service

Staffed expertise, fast turnaround, and a cut of every recovery, forever.

fees forever
Own

Own a scoped appeals agent

It drafts every letter and submits none of them. Fixed scope, fixed price, eval suite included.

the fourth option →
How it works

A denial-appeals pipeline, fixed scope, delivered in six weeks

01

Intake

Document processing reads your denial letters and EOBs (payer, denial code, service, dates) with no manual re-keying.

02

Retrieval

A RAG pipeline over the payer’s own policy documents and the patient’s chart context finds the medical-necessity criteria the denial ignored.

03

Draft

The agent writes the appeal letter, citing the payer’s policy language and the clinical evidence, in your letterhead format.

04

The gatehuman judgment

A clinician reviews and signs every letter before it goes anywhere. Enforced in code, not in a policy document. The system cannot submit. That constraint is permanent and non-configurable.

Scope is deliberately narrow: one specialty, your top five denial reasons, one EHR export format. That is what makes the price fixed and the timeline honest. Expansion is a follow-on, priced the same way, in the open.

The differentiator

The eval suite is the product

Every vendor in this market claims accuracy. None of them shows you the harness. My acceptance criteria are written into the contract as eval scores, and you keep the machinery that produces them.

eval suite · acceptance criteria ships with the build

A golden dataset

Built from your historical denials and their outcomes, not synthetic examples.

your data

Citation-accuracy tests

Every policy citation in a drafted letter is checked against the actual payer document. Letters that cite policies that don’t say what the letter claims fail the suite.

fails on misquote

Factual-grounding regression

Drafted clinical statements are verified against the chart context they came from.

regression

A parallel-run gate

The agent runs alongside your current process until it clears the thresholds on live denials. No cutover before the score clears.

blocks cutover

When I hand over, you own the eval harness, the golden dataset, the regression suite, and a runbook. You don’t have to trust my accuracy claim. You can re-run it: this quarter, next quarter, after every model update. An appeals agent without a measurable judgment gate is a liability with a nice demo.

The price

The price is $28,000. Here is exactly what it buys.

One price. No tiers, no “starting at,” no quote theater.

One public price
$28,000
one-time · 6 weeks + 30-day stabilization
Schedule a call
  • Discovery workshop: your denial taxonomy, your top five denial reasons, where the clinician gate sits
  • Denial-letter and EOB intake pipeline (IDP)
  • Payer-policy RAG pipeline for one specialty
  • Appeal-letter drafting agent with policy citations
  • Clinician review-and-sign workflow (the enforced gate)
  • Golden dataset built from your historical denials and appeal outcomes
  • Eval harness: citation accuracy, factual grounding, regression suite
  • Guardrails in code: no auto-submission, PHI-handling constraints, confidence-floor escalation
  • Parallel-run period against your current process, with published thresholds
  • Handover pack: runbook, eval documentation, your team trained to run the suite
What’s not in the number, said plainly: you pay your own LLM API and hosting costs, typically $200–$600/month at practice volume, on your accounts, visible to you. Optional monitoring retainer after stabilization: $1,500/month, cancel anytime. Neither is required to run what you own.
The field

Against the alternatives

This build Enterprise platform Appeals service Freelancer
Price $28,000, once, public Undisclosed; six-figure annual territory Contingency on every recovery, forever Hourly, open-ended
Timeline 6 weeks + 30-day stabilization Months of procurement first Fast, but never yours Unknown
Who owns it You: code, evals, golden set The vendor The vendor You, minus the proof
Proof it works Eval suite you can re-run Their accuracy claim Their track record None shipped
Clinician control Gate enforced in code Varies by configuration Outside your walls Whatever got built

A practice losing $150K a year to unworked denials recovers the build cost in the first quarter it starts appealing at volume. I won’t promise your overturn rate; that’s what the parallel run measures. I will show you the score before you cut over.

Delivery

How the six weeks run

  1. Week 1

    Discovery

    Map the denial workflow, pull historical denials and outcomes, define the top-five reasons in scope, and fix the acceptance thresholds in writing.

  2. Weeks 2–3

    Golden dataset + pipeline

    Label your historical denials and appeal outcomes; stand up intake and the payer-policy RAG.

  3. Weeks 4–5

    Drafting agent + gate

    Letter generation with citations, the clinician sign-off workflow, guardrails in code.

  4. Week 6

    Parallel run

    The agent drafts alongside your current process. We compare against the golden set and live clinician judgments. Cutover only when thresholds clear.

  5. Days 1–30 after

    Stabilization

    I watch the evals, fix drift, tune escalation. Then you own it outright.

FAQ

Straight answers

The questions every RCM lead and practice administrator asks before booking the call.

Does the AI submit appeals automatically?

No, and it can't be configured to. Every letter requires clinician sign-off, and the constraint lives in code. Denial-automation liability is the top reason healthcare AI projects die in legal review; this design is how the project survives it.

What exactly does the $28,000 include?

Discovery, intake, RAG, drafting agent, clinician gate, golden dataset, eval harness, parallel run, handover, and 30 days of stabilization. The only costs outside the number are your own API/hosting spend and the optional monitoring retainer.

How do you prove it works on our denials, not just in a demo?

The golden dataset is built from your historical denials and outcomes, and the contract's acceptance criteria are eval thresholds on that set. Then a parallel run tests it against live denials before anything cuts over. You are never asked to take an accuracy claim on faith.

Which specialties and EHRs do you support?

The build covers one specialty and one EHR export format; that's what keeps the price fixed. On the call we check whether your denial mix and export path fit. If they don't, I'll tell you, and you'll have lost 30 minutes, not a deposit.

Is this HIPAA compliant?

The architecture is BAA-ready: PHI-handling guardrails in code, minimum-necessary retrieval, and the option to run on infrastructure you control. Compliance obligations and the BAA itself are scoped in discovery and stated plainly, not hand-waved.

What happens after handover?

You own the code, the golden dataset, and the regression suite, plus a runbook your team is trained on. Models drift and payers rewrite policies, so re-run the suite on a schedule: yourselves, or through the $1,500/month monitoring retainer, which you can cancel anytime.

Why is the price public when every competitor quotes custom?

Because scoped work can be priced, and unscoped work shouldn't be sold. The scope on this page is genuinely fixed, so the price can be too. If discovery reveals your situation needs more, you'll get a new scope with a number on it before any work starts.

What teams say

What revenue-cycle teams say

“His vast knowledge of technologies and a natural problem-solving mindset consistently lead us through complex challenges with clarity and confidence.”
AM
Adil Multani
Senior Backend Developer
Next step

Appeals win. File them.

The overturn statistics haven’t changed for years: most denials that get challenged, fall. The bottleneck is drafting capacity, and drafting is exactly the work a machine should do while your clinician holds the pen that matters.

30 minutes · the price stays $28,000 · if it’s not a fit, I’ll say so

Ask AI about AI Denial & Appeals Agent