End-to-end operational intelligence project focused on analyzing process performance, operational efficiency, data quality, process complexity, exceptions, and workflow automation using Python, PostgreSQL, Power BI, Power Apps, and Power Automate.
The objective of this project is to analyze operational process data and transform raw event logs into actionable business insights.
The analysis focuses on:
- Process performance
- Process duration
- Process complexity
- Operational exceptions
- Escalations and rework
- Customer satisfaction
- Event transitions
- Data quality
- Operational investigation
- Workflow automation
- Python
- Pandas
- PostgreSQL
- SQL
- Power BI
- DAX
- Power Apps
- Power Automate
- GitHub
The process mining event log was analyzed and transformed using Python and Pandas.
Python was used to:
- Load and prepare the event log dataset
- Aggregate events at case level
- Calculate process duration
- Calculate event counts
- Identify escalations
- Identify rework and reopening events
- Calculate process complexity
- Classify operational processes
- Prepare analytical datasets for PostgreSQL and Power BI
The processed datasets are available in the data/ folder.
The processed data was stored and analyzed in PostgreSQL.
SQL was used to:
- Create and structure analytical tables
- Define data types and constraints
- Analyze operational performance
- Calculate process metrics
- Analyze event transitions
- Identify operational hotspots
- Support the Power BI analytical model
The SQL scripts used in the project are available in the sql/ folder.
Processes were classified into three operational categories based on exception and complexity indicators:
- Standard — regular processes with lower complexity
- Exception / Rework — processes involving operational exceptions or rework
- Complex — processes with higher operational complexity
The classification supports the identification of processes that require deeper operational investigation.
The dashboard was developed in Power BI and organized into three analytical pages:
- Executive Overview — provides a high-level view of process volume, average duration, customer satisfaction, process classification, and operational duration gaps.
- Process Intelligence — analyzes process behavior, duration, complexity distribution, event counts, process variants, and event transitions.
- Operational Investigation — identifies operational hotspots by issue type, report channel, priority, escalations, rejections, reopenings, and long-duration cases.
The analysis highlighted several relevant operational patterns:
- Standard processes represent the majority of the analyzed cases.
- Complex processes present higher average process duration than Standard processes.
- Exception / Rework and Complex processes show lower average customer satisfaction than Standard processes.
- Escalations represent an important operational event within the process.
- Process complexity varies across cases and can be used to support operational investigation.
- Long-duration cases provide opportunities for deeper investigation of operational inefficiencies.
- Missing resolver values are structurally associated with specific event types and were therefore retained as NULL values.
The project also includes a conceptual workflow demonstrating how operational investigation can be connected to workflow automation.
A conceptual Power Apps interface was designed to allow users to register cases for operational investigation directly from the analytical workflow.
A conceptual Power Automate workflow was designed to notify the investigation team when a case requires follow-up.
Note: The Power Apps and Power Automate interfaces are conceptual mockups created to demonstrate the intended workflow integration.
operational-process-intelligence/
│
├── data/
│ ├── process_analysis.csv
│ └── process_mining_event_log.zip
│
├── python/
│ └── 01_process_analysis.py
│
├── sql/
│ ├── 01_create_process_analysis.sql
│ └── 02_operational_analysis.sql
│
├── powerbi/
│
├── powerapps/
│ └── Case_Investigation_App.png
│
├── powerautomate/
│ └── Operational_Investigation_Notification.png
│
├── images/
│ ├── Executive_Overview.png
│ ├── Process_Intelligence.png
│ └── Operational_Investigation.png
│
├── docs/
│
├── LICENSE
└── README.md
Project Outcome
This project demonstrates an end-to-end operational intelligence workflow, connecting data preparation, process analytics, data quality, SQL analysis, business intelligence, operational investigation, and workflow automation.
The solution follows the flow:
Data → Python → Data Quality → PostgreSQL → Power BI → Investigation → Automation