This is what an AI agent reads. Same page, second kind of user.

1---
2title: "AI Reimbursement Engine for DME | Case Study"
3description: "How we built a generative AI engine that scores DME reimbursement likelihood against Medicare and Medicaid policy, before the device ships."
4url: https://appstream.studio/case-studies/ai-reimbursement-eligibility-engine
5markdown_url: https://appstream.studio/case-studies/ai-reimbursement-eligibility-engine.md
6---
7 
8[Home](https://appstream.studio/index.md) · [Work](https://appstream.studio/case-studies.md) Healthcare
9 
10Healthcare/ Engineer
11 
12# Score reimbursement likelihood before the device ships.
13 
14A DME startup was writing off revenue because nobody could predict which placements would get reimbursed. We built an AI engine that checks every case against Medicare and Medicaid policy before anything goes out the door.
15 
16Client
17 
18Healthcare Startup (DME)
19 
20Engagement
21 
22Ongoing
23 
24Focus
25 
26AI Agents + RAG
27 
28Compliance
29 
30HIPAA, Medicare, Medicaid
31 
32![AI reimbursement engine for durable medical equipment](https://appstream.studio/ai-reimbursment-eligibility-engine.png)
33 
3401
35 
36## The work
37 
38A healthcare startup in the durable medical equipment space, running patient programs that provide medical devices under Medicare and Medicaid coverage. Industry-wide, an estimated 20 to 50 percent of DME placed in the field fails reimbursement and gets written off.
39 
40What was getting in the way
41 
42- A meaningful share of devices shipped were never reimbursed. Equipment and effort turned into bad debt.
43- Eligibility logic was scattered across staff, binders, and tribal knowledge. No two people applied it the same way.
44- Checking reimbursement meant manually cross-referencing Medicare and Medicaid rules against patient records. It didn't scale.
45- There was no system to score how likely a case was to be paid, or flag what documentation was missing.
46- When payers denied claims, the team couldn't quickly build a structured argument for medical necessity.
47 
4802
49 
50## What changed
51 
52Train AI on real policy. Feed it real patient data.
53 
54We took the specific Medicare and Medicaid policies relevant to the client's DME portfolio and structured them into a knowledge layer the AI could reason over. Organized by device type, program, conditions, thresholds.
55 
5603
57 
58## What we built
59 
60Then we connected it to live patient data from their EMR via HL7 and FHIR. Diagnoses, clinical history, program enrollment, device data. All normalized and governed for PHI compliance.
61 
62The engine matches each case against the relevant rules, produces a reimbursement score, and explains which criteria are met, which are borderline, and what's missing. Staff see the answer before the device ships. If a claim gets questioned later, the documentation is already there.
63 
64Under the hood, the engine runs on Anthropic's Claude models: Opus 3 for the heavy policy reasoning and Sonnet for faster, high-volume scoring. Our team built it with Claude Code, which let a small crew move through the policy structuring, integration work, and testing much faster.
65 
66AI Agents RAG Pipelines System Integration HL7/FHIR Integration HIPAA Audited
67 
6804
69 
70## How it works
71 
72- AI engine trained on the Medicare and Medicaid rules that actually govern this client's DME categories
73- Live patient and clinical data flowing from the EMR via HL7 and FHIR into the scoring engine
74- Per-case reimbursement scores with plain explanations of what's satisfied and what's not
75- High-risk cases and missing documentation flagged before a device goes out
76- Justification material for audits and appeals, generated from the same scoring run
77- A platform built to extend as new device types, programs, and payer rules come in
78 
7905
80 
81## Business impact
82 
83industry write-off rate
84 
8520-50%
86 
87Of DME placed industry-wide fails reimbursement. That's the baseline we're fixing.
88 
89scoring
90 
91Real-time
92 
93Per-patient reimbursement likelihood, checked before the device ships
94 
95documentation
96 
97Audit-ready
98 
99Justification tied to specific Medicare and Medicaid rules, generated automatically
100 
101> “We used to guess. Now we know before we ship whether a placement will get paid.”
102 
103Founder , DME Healthcare Startup
104 
10506
106 
107## What’s next
108 
109Stop writing off revenue you should be collecting
110 
111If reimbursement decisions at your organization depend on tribal knowledge and manual policy review, we should talk.
112 
113[Talk to an AI engineer](https://appstream.studio/schedule-a-discovery-call.md) · [More work](https://appstream.studio/case-studies.md)
114 
115Team
116 
117- Justin Tannenbaum Solutions Architect
118- Lukasz Chmielewski Lead Engineer
119- Mikolaj Kaminski Engineering
120- Daniel Bukala Product Manager
121 
122Stack
123 
124.NET · C# · Azure · Semantic Kernel · Claude Opus 3 · Claude Sonnet · Claude Code · HL7 · FHIR
125 
126Related
127 
128- [Engineer](https://appstream.studio/solutions/engineer.md)
129- [Agent Engineering](https://appstream.studio/solutions/agent-engineering.md)
130- [Agent Knowledge Layer](https://appstream.studio/solutions/agent-knowledge-layer.md)
131- [Agent Operations](https://appstream.studio/solutions/agent-operations.md)
132- [Healthcare](https://appstream.studio/industries/healthcare.md)
133 
134More work
135 
136## Keep reading.
137 
138[All work](https://appstream.studio/case-studies.md)
139 
140- [Healthcare/ Engineer · Operate A scheduling engine on a legacy EMR books 250,000 appointments a year. Read the case study](https://appstream.studio/case-studies/healthcare-scheduling-transformation.md)
141- [Proptech/ Engineer A 50-state buy-side marketplace, with search that understands intent. Read the case study](https://appstream.studio/case-studies/buy-side-marketplace-agentic-search.md)
142- [Private Equity/ Engineer · Operate One AI assistant answers HR questions across every acquired company. Read the case study](https://appstream.studio/case-studies/enterprise-genai-people-operations.md)
143 
144---
145 
146## AppStream Studio
147 
148AppStream turns frontier AI capabilities into production systems that perform real work. Talk to an AI engineer: https://appstream.studio/schedule-a-discovery-call · Email: contact@appstream.studio
149 
150- [Home](https://appstream.studio/index.md)
151- [Solutions](https://appstream.studio/solutions.md)
152- [Advise](https://appstream.studio/solutions/advise.md)
153- [Engineer](https://appstream.studio/solutions/engineer.md)
154- [Operate](https://appstream.studio/solutions/operate.md)
155- [Govern](https://appstream.studio/solutions/govern.md)
156- [How we work](https://appstream.studio/how-we-work.md)
157- [Work (case studies)](https://appstream.studio/case-studies.md)
158- [Industries](https://appstream.studio/industries.md)
159- [Insights](https://appstream.studio/blog.md)
160- [Partners](https://appstream.studio/partners.md)
161- [About](https://appstream.studio/about-us.md)
162- [Contact](https://appstream.studio/contact.md)
163