Coding / HIPPS Classification
Automate deterministic coding and HIPPS/PDPM classification, with human review where judgment is required.
What the team is doing right now.
Classification and coding pull data across clinical and therapy systems and re-key it; the deterministic parts are slow, the judgment parts are bottlenecked, and accuracy depends on who happens to be at the desk.
Scope.
- 01PDPM and HIPPS scoring support from clinical and therapy data
- 02Diagnosis ingestion and normalization
- 03Coding assistance with human review on edge cases
- 04Surfacing of evidence supporting each code
Inputs → steps → HITL gate → outputs.
Standard systems and EDI sets only.
Primary value vector: Capacity.
Move the deterministic majority of this work to the agent so the people who used to do it can work the work that actually needs them.
No RCM Kit performance number is stated on this page. Industry-benchmark ranges shown in the home page problem framing and ROI Estimator are illustrative only.
Core SNF strength — PDPM components, HIPPS coding, MDS-driven classification.
SNF is our flagship proof of depth, not the limit of scope. The same agent extends across broader healthcare RCM.
Honest answers to the obvious objections.
We've heard most of these. Here's where we stand.
How do you handle accuracy and hallucination risk?
Every output with regulatory or financial exposure is reviewed by a person before it leaves the system. Agents must clear evaluation gates before they go live and to stay live, and autonomy expands only as measured confidence accrues for that agent on that workflow.
How is data and PHI handled?
Agents handle protected health information per HIPAA-aligned practices: data segregation, least-privilege access, encryption in transit and at rest, and complete audit logging. Security and compliance are part of the architecture, not bolted on. See the Security page for full detail.
How heavy is the integration lift?
Agents execute the workflow inside your existing systems — EHR/EMR, billing, clearinghouses, payer channels. Deployment is phased, agent by agent, often shadow-running before any autonomy. You are not adopting a new operating system; you are deploying a workforce inside the one you have.
What's available now vs. on the roadmap?
Payment Posting is available today. Accounts Receivable & Follow-Up is in active development with an early-access cohort forming. The other 16 agents are on the roadmap — the waitlist is how you get in early and influence priority.
How does pricing work?
There is no public pricing yet. Early-access conversations are how we shape pricing with the first cohort. Join the waitlist and we will be in touch.
How is this different from RCM AI tools that failed before?
Prior generations were sales-led, horizontal, shallow on domain, and uncontrolled — autonomy without a safety story. RCM Kit is specialized (SNF-flagship, extending across healthcare RCM), deeply integrated into the systems where the work actually happens, and control-first: human-in-the-loop, evaluation gates, full audit. Different architecture, different result.
Notify me when the Coding / HIPPS Classification agent is live.
We're rolling out the suite agent by agent. The early-access cohort is small by design — early onboarding, priority on roadmap influence, and a direct line to the team building it.
- → Onboarding before general availability
- → Roadmap influence on which agents ship next
- → Direct access to the build team — no SDR layer