Agent skills for healthcare and life sciences: genomics, imaging, claims, drug discovery, and more. Works with Amazon Quick, Kiro, Amazon AgentCore, AWS Strands SDK, Claude Code, Codex, and any Agent Skills-compatible platform.

A library for stance detection, claim extraction, and more

Python ETL pipeline automating healthcare claims auto-adjudication reporting — ingestion, validation, metrics, Excel reports, and email delivery across 5 business segments.

A fact-checking system that verifies user claims with evidence-backed explanations using web searches and AI, delivering clear, transparent verdicts through a simple interface.

Multi-modal insurance claims review agent built for HackerRank Orchestrate Hackathon June 2026

End-to-end healthcare claims analytics: Synthea → DuckDB → dbt (sources/staging/intermediate/marts) → Streamlit + Plotly ops UI. Surfaces $2.6M outstanding across 14K claims with a ranked priority queue, claim-level drill-down, and explainable "why flagged" rules. Modern data stack portfolio build.

AI-powered Python tool using LLMs to automate insurance form autofill from unstructured documents. Boosts efficiency for claims, underwriting, and policy processing.

Pre-submission claim scrubbing and denial-risk engine for hospital revenue cycle. 11 configurable payer edits, calibrated risk scoring, expected-loss ranked review workqueues, and Power BI-ready SQL views. FastAPI, PostgreSQL, scikit-learn.

ClaimFlow — Autopilot Agent for Claims Processing | Track 4: Autopilot Agent | Global AI Hackathon with Qwen Cloud

Reference architecture for governed agentic AI on AWS. Purpose-bound Cedar authorization at column level, Iceberg snapshot-pinned queries for replayable audit trails, and confidence-gated document extraction with Bedrock Data Automation and AgentCore, demonstrated on insurance claims processing with an EU AI Act mapping.

MediRegs Advocate: WK Code Games Q1 entry. A judgment engine on the MediRegs regulatory corpus: claims checked pre-bill, appeals drafted with version-proof citations, audit exposure mapped from public data. Clean-room prototype, public CMS sources only.

Multi-agent insurance claims processing - extraction with citation grounding, deterministic policy validation, and audit-ready adjudication packets

Multi-modal damage-claim evidence-review agent: two-call Claude vision pipeline (claim-blind perception + claim-aware verdict), deterministic coupling rules, and structured Pydantic output over car/laptop/package claims. Built for the June 2026 HackerRank Orchestrate hackathon (165th of 1,773).

Pre-submission EDI 837 validation across BCBS, Aetna, Cigna, Humana, and Medicare. Flags $1.16M in revenue at risk across 1,200 synthetic claims per run

Mines 837/835-shaped hospital claims for preventable denials and prices revenue leakage by CARC code, payer, and department. 99.5% precision and 100% recall measured against labeled synthetic ground truth, 126K claims/sec on 1 vCPU, with a full Power BI semantic model included.

AI-Powered Intelligent Claims Adjudication Layer for HealthFlow HCX Platform — Eligibility, Medical Coding, Fraud Detection & Medical Necessity agents built on LangGraph, LiteLLM, and ChromaDB RAG

Autonomous FNOL insurance claims processing agent that extracts key claim fields from TXT/PDF documents, detects missing or inconsistent information, classifies claims, and recommends routing decisions with JSON output and plain-English reasoning.