Problem: Auto-Adjudication (AA) reporting is manual, slow, and decision-heavy.
This repository showcases a Streamlit-based UI that automates claims data processing and turns a fragmented AA reporting workflow into a streamlined, end-to-end pipeline—from raw mainframe data to shareable, email-ready reports.
The existing AA reporting process is fragmented and labor-intensive:
- Manual data handling — Teams extract JCL data into Excel
- Slow processing — Requires cleanup, validation, and human decision-making
- Error-prone — Inconsistent schemas, manual transformations, and duplicate efforts
- Poor visibility — Stakeholders wait for Excel reports and email handoffs
- Siloed storage — Results scattered across SharePoint, email archives, and local drives
JCL/Mainframe Extract → Manual Excel Cleanup → Validation → Email → SharePoint Storage
❌ Pain points:
- Time-consuming manual steps
- High risk of human error
- Delayed insights and reporting
- No audit trail or consistency
CSV/TXT Upload → Ingest → Validate → Transform → Aggregate → Export → Email Preview
✅ Key improvements:
- Centralized Streamlit UI — Single point for file upload and pipeline execution
- Automated data pipeline — Ingest, validate, transform, and aggregate in seconds
- Intelligent AA transformation — Business logic to process claims and compute metrics
- Real-time dashboards — View overall AA rate, segment breakdowns, and key metrics
- Email automation — Generate HTML-ready reports with one click
- Historical tracking — Automatically appends daily AA metrics to historical datasets (
westmarket.xlsx) for trend analysis - Configuration-driven — Easy customization via
config.yamlfor recipients, paths, and templates
| Metric | Improvement |
|---|---|
| Processing Time | ⏱️ From hours to seconds |
| Manual Effort | 📉 ~80% reduction in Excel work |
| Data Quality | ✅ Automated validation eliminates human error |
| Stakeholder Visibility | 📊 Real-time metrics and summaries |
| Consistency | 🔄 Repeatable, audit-ready pipeline |
Include screenshots or a GIF here to show the UI flow and output.
- Screenshot: input upload screen
- Screenshot: AA metric summary
- GIF: report preview and email-ready output
- Python 3.8+
- pip or conda package manager
git clone https://github.com/muzammil-13/claims-reporting-ui.git
cd claims-reporting-uipython -m venv venv
# Activate virtual environment
# On Windows:
venv\Scripts\activate
# On macOS/Linux:
source venv/bin/activatepip install -r requirements.txtEdit config.yaml to customize:
- Input and output directories
- Email recipients and sender
- Report subject templates
streamlit run app.pyThe app will open in your browser at http://localhost:8501
claims-reporting-ui/
│
├── app.py # 🎨 Streamlit UI entry point
├── config.yaml # ⚙️ Configuration (paths, email, templates)
│
├── pipeline/ # 🔄 Core data pipeline
│ ├── ingest.py # Load CSV/TXT from mainframe exports
│ ├── validate.py # Schema validation and data quality checks
│ ├── transform.py # AA business logic and claims processing
│ ├── aggregate.py # Compute metrics (AA rate, LOB summaries)
│ └── export.py # Generate Excel reports (YTD) and update historical datasets (Westmarket)
│
├── automation/ # 📬 Reporting automation
│ └── email.py # Generate HTML email content (Outlook/SMTP ready)
│
├── data/
│ ├── input/ # Raw mainframe extracts (CSV/TXT)
│ └── output/ # Generated reports and exports
│
├── requirements.txt # Python dependencies
└── README.md # This file
| Component | Technology |
|---|---|
| UI Framework | Streamlit 1.20+ |
| Data Processing | Pandas 1.5+ |
| Configuration | PyYAML 6.0 |
| Excel Support | OpenPyXL 3.1+ |
| Formatting | Tabulate 0.9+ |
- Upload — User uploads raw claims data (CSV/TXT from mainframe)
- Preview — Pipeline shows data preview for validation
- Execute — Click "Run Pipeline" to trigger automated workflow
- Process —
- ✅ Data Ingested (load into DataFrame)
- ✅ Data Validated (schema checks, required fields)
- ✅ Data Transformed (apply AA business logic)
- ✅ Metrics Aggregated (calculate AA rates, LOB summaries)
- ✅ Reports Generated (export YTD reports with Raw Data & Segment Summary sheets, and update historical trends)
- Review — See AA rate, segment summaries, and key metrics
- Share — Generate HTML email preview or send directly
paths:
input_dir: "data/input" # Where uploads are stored
output_dir: "data/output" # Where reports are generated
email:
sender: "pipeline@example.com"
recipients: ["stakeholders@example.com"]
subject_template: "Daily Healthcare Claims AA Report - {date}"- 🔗 SharePoint API integration — Automated report publishing to SharePoint
- 📈 Interactive dashboards — Charts for AA trends over time
- ⏰ Scheduler support — Cron / Apache Airflow integration for automated runs
- 🔐 Backend authentication — Real database integration for secure data access
- 📄 PDF export — Direct PDF generation for archival
- 🧪 Unit testing — Comprehensive test coverage for pipeline stages
- 🐳 Containerization — Docker support for easy deployment
- 📊 Advanced analytics — Drill-down analysis by line of business, claim type, etc.
This project demonstrates:
- Enterprise workflow automation — Translating manual processes into scalable pipelines
- Data engineering in production — Ingest → validate → transform → aggregate → export
- Rapid prototyping with Streamlit — Building functional UIs without frontend overhead
- Configuration-driven design — Easy customization without code changes
- Error handling & validation — Robust schema validation and graceful failure modes
- Business logic implementation — Real-world AA calculation rules and metrics
- Built during: Internship at IBM Consulting Client Innovation Center
- Domain: Healthcare Claims Processing & Auto-Adjudication (AA) Reporting
- Inspired by: Real-world challenges in enterprise claims processing workflows
- Status: 🚀 Active and ready for customization
| Task | Command |
|---|---|
| Install dependencies | pip install -r requirements.txt |
| Run the app | streamlit run app.py |
| View config options | cat config.yaml |
| Access UI | http://localhost:8501 |
If you're working on similar automation, data pipeline, or healthcare integration projects, feel free to reach out!
⭐ If you found this useful, consider giving it a star!
Made with ❤️ for healthcare data automation.