Generates SDTM datasets and SAS programs for 3 domains (DM, AE, LB) with 5 subjects. Click to see the 6-agent pipeline in action.
| Job Requirement | This Project |
|---|---|
| Design AI agent apps for SDTM automation | 6-agent pipeline with MCP orchestration |
| SDTM dataset creation, quality checks | 8 CDISC-compliant domains with QC |
| Generate SAS programs | Full SAS DATA step + SORT + LABEL |
| Apply CDISC SDTM, CDASH | SDTMIG 3.4 spec engine (100+ rules) |
| LLM frameworks (LangChain, CrewAI) | LangChain-ready, MCP protocol |
| Azure OpenAI, AWS Bedrock | Drop-in LLM provider support |
Upload your own clinical CSV data and this agent will apply SDTM mappings, generate the compliant dataset, and produce the SAS program. See sample formats.
Below are the SDTM variable mapping rules for each domain. Click a domain to expand its specification. Each variable shows its mapping type: direct decode hardcoded computed
SDTM Automation Agent — Built for the AI Agent Developer — SDTM Automation role at Redbock/NES Fircroft ($80-105/hr, Remote US, 12-month contract).
This portfolio project demonstrates the exact combination required: Statistical Programming + CDISC SDTM expertise + AI agent development + Python + SAS code generation + MCP protocol.