📊 FlowSense AI — Smart Workflow Automation Advisor FlowSense AI is a modular, AI-powered advisor designed to help enterprises simulate and optimize workflow automation strategies. Built for Ricoh’s digital workplace ecosystem, it combines process mining, ROI simulation, and explainable AI to guide stakeholders toward measurable, compliant automation outcomes.
🚀 Purpose Enable enterprise teams to:
Analyze system logs for automation opportunities
Simulate ROI of automation strategies
Integrate with Ricoh’s document management systems
Ensure transparency with explainable AI insights
🧠 Key Features 🔍 AI-Driven Process Mining: Extract patterns from system logs to identify automation candidates
📈 ROI Simulation Engine: Estimate time savings and efficiency gains from proposed strategies
🔗 Ricoh Integration: Seamless connection to document workflows and enterprise systems
🧾 Explainable AI Layer: SHAP-based transparency for stakeholder trust and compliance
☁️ Azure Functions: Trigger predictions and simulations via cloud-based endpoints
📊 Power BI Export: Visualize insights in enterprise dashboards
🧰 Tech Stack Layer Tools Used AI Engine PyTorch UI & Simulation Streamlit BI Integration Power BI Cloud Functions Azure Functions Explainability SHAP 📁 Project Structure bash FlowSense-AI/ │ ├── app.py # Streamlit UI for simulation and prediction ├── workflow_model.py # PyTorch model for workflow classification ├── explain.py # SHAP-based explainability layer ├── export_to_powerbi.py # Data export for Power BI integration ├── function_app.py # Azure Function for cloud-based triggers ├── README.md # Project overview and instructions 🧪 Getting Started Clone the repo:
bash git clone https://github.com/MampotjeMabusela/FlowSense-AI.git cd FlowSense-AI Install dependencies:
bash pip install -r requirements.txt Run the Streamlit app:
bash streamlit run app.py Upload a system log CSV and view automation recommendations.
🔐 Compliance & Ethics FlowSense AI is designed with POPIA/GDPR in mind:
Logs are anonymized before processing
Role-based access control via Azure Identity
Audit trail maintained for all predictions
📌 Use Cases Legal tech teams evaluating smart contract workflows
IT departments optimizing document routing
Compliance officers validating automation transparency
Strategy consultants simulating ROI for digital transformation
🤝 Contributing We welcome contributions that enhance:
Model accuracy and explainability
UI/UX improvements
Integration with additional Ricoh services
📬 Contact Built by Mampotje Mabusela — aspiring AI specialist and legal tech innovator. For consulting, collaboration, or feedback, feel free to reach out via GitHub or LinkedIn.