EDI mapping that thinks for you
CData Arc embeds AI directly into the transformation layer—automatically suggesting field mappings, translating between formats, and making complex EDI configuration accessible to any team member. Onboard trading partners in hours, not weeks.
AI-assisted mapping · AI runtime connector · OpenAI, Gemini & local LLMs
EDI mapping has always been the hard part
Every new trading partner brings custom field structures, format quirks, and weeks of back-and-forth configuration. The bottleneck isn’t the protocol—it’s the mapping.
Weeks per partner onboarding
Manual field-by-field mapping between EDI transaction sets and internal formats can take 2–6 weeks per partner, consuming specialist time and delaying go-live.
Expertise bottleneck
EDI mapping requires deep knowledge of X12, EDIFACT, and partner-specific variations. A single expert’s availability dictates your entire onboarding pipeline.
Error-prone configuration
Manually mapping dozens of fields across complex schemas introduces silent errors that only surface during partner testing—causing rejections and delays.
Scale without scale
Adding 10 new retail partners shouldn’t require 10x more expert hours. Traditional mapping tools don’t get faster with experience—AI does.
Intelligent field mapping built into the XML designer
Arc’s AI mapping assistant analyzes the structure and semantics of your source and target schemas, then suggests transformation logic—dramatically reducing the time and expertise required to build accurate maps.
Semantic field suggestions
The AI understands what fields mean, not just their names. It recognizes that BEG03 in an X12 850 maps to PurchaseOrderID in your ERP schema—and suggests the link with a confidence score.
Configurable AI models
Connect to OpenAI (GPT-4), Google Gemini, or locally hosted Ollama LLMs. Use cloud intelligence for speed or local models for air-gapped environments where data never leaves your network.
One-click apply
Review AI suggestions inline within the XML designer, accept or adjust them with a click, then validate. What used to take days of expert configuration now takes minutes of review.
From schema to working map in minutes
The AI mapping workflow integrates seamlessly into Arc’s existing XML designer—no separate tool, no context switching.
1
Load your schemas
Import your source EDI schema (X12 850, EDIFACT ORDERS, etc.) and your target format (ERP XML, JSON, flat file).
2
Request AI suggestions
Arc sends the schema structures to your configured AI model. The model analyzes field names, data types, and EDI semantics to generate mapping recommendations.
3
Review & accept
Each suggestion shows a confidence score. Accept high-confidence mappings in bulk, manually review edge cases, and apply your full map in one action.
4
Test & deploy
Validate the map against real partner test data in Arc’s built-in testing environment, then deploy to your live flow with zero downtime.
Embed AI intelligence into any data flow
Beyond mapping assistance, Arc’s AI Connector brings runtime AI processing into your integration workflows—as a trigger, transformer, or terminal step in any flow.
Multi-provider support
Connect to OpenAI, Google Gemini, Ollama, or any REST-compatible AI model. Switch providers without changing your flow logic—just update the connection configuration.
Dynamic ArcScript prompts
Use ArcScript variables to build context-aware prompts that pull values from your message data. Process EDI documents with prompts tailored to the specific transaction, partner, or content.
Structured output generation
Transform unstructured text from partner communications, PDF invoices, or email attachments into structured EDI or XML that flows directly into your downstream systems.
Real-time translation
Translate partner messages, product descriptions, or compliance text across languages in-flight—without external preprocessing steps or manual intervention.
Your AI model, your terms
Arc’s AI capabilities work with any provider you choose—cloud or local, managed or self-hosted.
OpenAI / Gemini
Connect to GPT-4 or Google Gemini for best-in-class suggestion accuracy. Ideal for teams that already use cloud AI services and want maximum mapping quality.
Ollama (local LLMs)
Run open-weight models like Llama 3 or Mistral on your own infrastructure via Ollama. Sensitive EDI data—POs, invoices, healthcare transactions—never leaves your network.
Any REST-compatible model
Arc’s AI Connector speaks to any model with a REST API. Bring your own enterprise AI deployment, internal fine-tuned model, or future provider without changing your Arc configuration.
What changes when mapping takes minutes
AI-assisted mapping doesn’t just save time on individual maps—it changes the economics of partner onboarding entirely.
Partner onboarding time
Hours
vs. 2–6 weeks of manual mapping and testing
Mapping speed improvement
90% faster
from schema load to validated, deployable map
EDI expertise required
Drastically reduced
AI handles the field semantics; your team handles the business logic
Partners per expert per quarter
10× more
same team, same budget, dramatically higher throughput
Where AI mapping delivers the most value
Retail supplier onboarding
Onboard new retail trading partners (Walmart, Target, Amazon, Costco) with AI-suggested maps for 850/855/856/810/997 transaction sets. Get to first live transaction faster than any competitor.
ERP migration & replatforming
Remapping all partner integrations for a new ERP? AI dramatically accelerates schema migration by suggesting mappings between your old and new data models based on field semantics.
Healthcare data exchange
Map HIPAA 837/835/270/271 transactions to internal formats with AI assistance. Use local Ollama models to ensure PHI never leaves your environment during the mapping process.
Unstructured-to-EDI automation
Use the AI Connector to extract structured order data from PDF purchase orders, email attachments, or faxed documents—converting unstructured input to valid EDI in a single flow.