What We're Announcing
The Redix MCP Server is now available. It exposes healthcare data conversion tools through the Model Context Protocol (MCP), the open standard for connecting AI agents to external tools and data sources.
This means AI agents like Claude, Cursor, and any MCP-compatible client can now validate HIPAA X12 transactions, convert between healthcare data formats, load data to databases, and generate PDF claim forms - all through natural language.
Why it matters: Healthcare developers spend weeks writing format conversion code. With the Redix MCP Server, an AI agent can perform these conversions in seconds - no API expertise required.
What Is MCP?
The Model Context Protocol is an open standard created by Anthropic that lets AI models interact with external tools. Think of it as a universal adapter: instead of teaching every AI model how to call every API, MCP provides a single standard interface.
When you connect the Redix MCP Server to your AI client, the agent automatically discovers all the healthcare tools and can use them in conversation. You say "validate this 837 claim" and the agent calls the right tool, parses the result, and explains what it found.
What AI Agents Can Do
With the Redix MCP Server connected, your AI agent can handle multi-step healthcare data workflows in a single conversation:
Validate
"Check this 837P claim for errors" - validates structure, segments, and data elements against HIPAA 5010 rules.
Convert
"Convert this X12 835 to FHIR R4" - transforms between HIPAA X12, FHIR, HL7, CDA, RMap, and database formats.
Generate
"Create a CMS-1500 PDF from this claim" - generates print-ready PDF claim forms (CMS-1500, UB-04, ADA J400).
Chain
"Validate this claim, convert to FHIR, and generate a PDF" - the agent chains tools automatically across one conversation.
The Tools
| Tool | Description |
|---|---|
validate_x12 | Validate HIPAA X12 transactions against 5010 implementation guides |
convert_x12_to_fhir | Convert HIPAA X12 to a FHIR R4 Bundle (with Provenance for the 278) |
convert_x12_to_rmap | Convert HIPAA X12 to RMap intermediate format |
convert_rmap_to_x12 | Convert RMap back to HIPAA X12 transaction |
convert_x12_to_database | Load HIPAA X12 data into a relational database |
generate_x12_from_database | Generate HIPAA X12 transaction from database records |
convert_hl7_to_fhir | Convert HL7 v2.x messages to FHIR R4 |
convert_cda_to_fhir | Convert CDA/C-CDA documents to FHIR R4 |
convert_fhir_to_rmap | Convert FHIR R4 Bundle to RMap format |
generate_claim_pdf | Generate CMS-1500, UB-04, or ADA J400 PDF from claim data |
list_supported_formats | List all supported formats and transaction types |
The Redix Advantage: IFD/OFD Flexibility
The MCP Server and REST API are fixed wrappers - they provide consistent interfaces for AI agents and applications. But underneath, the Redix AnyToAny engine is fully configurable.
Nothing Is Hardcoded
- IFD (Input Format Definition) - flat files that define validation rules for each transaction type. You can modify segment requirements, data element constraints, and code set validations without touching code.
- OFD (Output Format Definition) - a proprietary language similar to Python/VB that defines conversion logic. Field mappings, conditional transforms, and output formatting are all configurable.
- Standard maps included - pre-defined maps cover the standard HIPAA transactions. Redix consultation services add each organization's specific payer or provider requirements.
This means the same MCP tools that work with standard transactions today can be adapted to your organization's custom requirements - without waiting for a software release.
How to Connect
Claude Desktop (Local)
Add to your Claude Desktop configuration file:
claude_desktop_config.json{
"mcpServers": {
"redix-anytoany": {
"command": "python3.11",
"args": ["/path/to/redix-mcp-server/server.py"],
"env": {
"REDIX_API_URL": "https://demo.redix.com",
"REDIX_API_KEY": "your-api-key"
}
}
}
}Claude Code / Cursor (Local)
.mcp.json{
"mcpServers": {
"redix-anytoany": {
"command": "python3.11",
"args": ["/path/to/redix-mcp-server/server.py"],
"env": {
"REDIX_API_URL": "https://demo.redix.com",
"REDIX_API_KEY": "your-api-key"
}
}
}
}Remote SSE (Any MCP Client)
For remote access without local installation, connect via Server-Sent Events:
SSE Endpointhttps://demo.redix.com/sseThe SSE transport works with any MCP-compatible client. No local Python installation required.
Validation Gates
The Redix MCP Server runs automated checks on the way in and on the way out:
Input validation
Validates incoming data against HIPAA 5010 implementation guides before conversion. Catches structural errors, missing segments, and invalid code values.
Output checks
Checks converted output against the target format's structure before it is returned. A structural check is not payer acceptance or implementation-guide certification.
Try It Now
Three ways to explore the Redix MCP Server today:
- MCP Inspector - Interactive browser tool to test the MCP tools directly. No installation required.
- Developer Portal - Full documentation, configuration examples, and tool reference.
- Live Sandbox - Test conversions with sample data through the web interface.
MCP Inspector: Visit demo.redix.com:6275 to interactively test every tool. Select a tool, provide input, and see results in real time - directly in your browser.
Deployment Options
| Option | Transport | Best For |
|---|---|---|
| Local (STDIO) | Subprocess | Development, testing, air-gapped environments |
| Remote (SSE) | Server-Sent Events | Evaluation on demo.redix.com with sample data; team access when you host it yourself |
| On-Premises | STDIO or SSE | Organizations where PHI cannot leave the network |
All deployment options use the same tools with the same Redix C-engine underneath; the MCP server is a thin wrapper around the engine. The public demo is for synthetic data only, never PHI.
Your AI agent. Your healthcare data. Connected.
Frequently Asked Questions
What is the Redix MCP Server?
The Redix MCP Server exposes healthcare data conversion tools via the Model Context Protocol (MCP). AI agents like Claude, Cursor, and other MCP-compatible clients can validate X12 transactions, convert between HIPAA, FHIR, HL7, CDA, and RMap formats, load data to databases, and generate PDF claim forms - all through natural language conversation.
How do I connect an AI agent to the Redix MCP Server?
For local use, add the server to your Claude Desktop or Cursor configuration as a STDIO subprocess pointing to the server.py file. For remote use, connect via Server-Sent Events (SSE) at https://demo.redix.com/sse. See the Redix Developer Portal for full configuration instructions.
Is the Redix conversion engine hardcoded?
No. The Redix engine uses IFD (Input Format Definition) flat files for validation and OFD (a proprietary language similar to Python/VB) for conversion. Nothing is hardcoded. Pre-defined maps cover the standard transactions. Consultation services complete the remaining customization for each organization's specific needs.