📞 DPDzero Data Ops Assignment

This project simulates a real-world DataOps pipeline for processing daily loan collection call data and generating agent performance summaries.

---

🎯 Objective

Build a simplified end-to-end pipeline that:

• Ingests raw daily CSV dumps (call logs, agent roster, login summary)
• Validates and merges the data
• Computes metrics like connect rate, presence, and call durations
• Outputs a CSV report and Slack-style summary message
• Optional: UI to upload files and interact with the report

---

🌐 Live App

👉 Open Streamlit App

---

🗂️ Folder Structure

dpdzero_assignment/
├── data/ # Input CSV files
├── output/ # Final report gets saved here
├── src/ # Modular pipeline scripts
│ ├── ingest.py
│ ├── validate.py
│ ├── merge.py
│ ├── features.py
│ ├── report.py
│ └── logger.py
├── main.py # CLI entry point
├── streamlit_app.py # Streamlit UI app
├── requirements.txt # Python dependencies
├── .gitignore
└── README.md

---

🚀 How to Run

1. Install Dependencies:

pip install -r requirements.txt

2. Run the pipeline with default files and date:

python main.py --date 2025-04-28

3. Optional CLI arguments:

--call_logs        Path to call_logs.csv
--agent_roster     Path to agent_roster.csv
--disposition      Path to disposition_summary.csv
--date             Date to generate summary for (YYYY-MM-DD)

4. StreamLit UI:

streamlit run streamlit_app.py

Upload your files, pick a date, and generate summary.

---

🧠 Metrics Computed (Per Agent Per Date)

• 📞 Total Calls Made
• 🔁 Unique Loans Contacted
• ✅ Connect Rate (% completed calls)
• ⏱️ Average Call Duration (in minutes)
• 👤 Presence (based on login time)

---

📦 Output

• output/agent_performance_summary.csv
• Slack-style summary printed to terminal or UI:

📊 Agent Summary for 2025-04-28
🏆 Top Performer: Ravi Sharma (98% connect rate)
👥 Total Active Agents: 45
⏱️ Average Duration: 6.5 min

---

✨ Features

✅ Modular code with reusable pipeline functions ✅ CLI and Web UI options ✅ Logging with timestamps ✅ Upload, run, and download from browser ✅ Graceful handling of missing data

---

🧑‍💻 Author

This project was developed as part of a hiring assignment by DPDzero.

🔧 Built With

• Python
• Pandas
• Streamlit
• argparse, logging