I design and build practical automation systems that improve business workflows, operational visibility, data handling, and process efficiency.
My current professional focus is manufacturing workflow automation — developing systems that connect operational processes, reduce repetitive manual work, improve information flow between departments, and provide better visibility for business teams.
I work primarily with Python, Django, n8n, APIs, databases, and AI-enabled components where they provide clear operational value.
- Manufacturing workflow automation
- Business process automation
- n8n workflow orchestration
- Python and Django development
- API and third-party system integrations
- Operational databases and structured data workflows
- Procurement and purchasing workflows
- Inventory and stores management
- Production planning and control
- Work-in-progress and shop-floor visibility
- Quality control and rework workflows
- Maintenance operations
- Packing, dispatch, and logistics workflows
- Operational dashboards and management visibility
- AI-assisted document processing, classification, summarization, and decision support
My approach is to use automation as the foundation and apply AI selectively where it improves the reliability, speed, or usability of a business process.
My first six portfolio projects established a foundation in customer-support automation, structured data retrieval, full-stack Django development, workflow management, operational dashboards, testing, and human-in-the-loop systems.
These projects use fictional businesses and synthetic data for portfolio and development purposes.
My most comprehensive completed portfolio system to date.
This project demonstrates a complete ecommerce customer-support operation rather than a standalone chatbot.
The system combines:
- Customer-facing support assistant
- Approved FAQ and policy retrieval
- PDF knowledge retrieval
- 150-product structured catalogue
- Product discovery
- Secure mock order verification
- Lead capture
- Human-support escalation
- Unanswered-question management
- Conversation tracking
- Business-data management
- Authenticated staff operations dashboard
- Role-based authorization
- Privacy-aware workflow controls
- Safe fallback behavior
- Extensive automated regression testing
The system follows an approved-source-first architecture, ensuring that customer-facing information comes from controlled business data while sensitive actions remain under explicit workflow and human oversight.
451 / 451 automated tests passed
Django system check: 0 issues
Key areas: workflow automation, full-stack Django systems, knowledge retrieval, operational dashboards, human-in-the-loop automation, privacy-aware workflows, regression testing
Tech: Python, Django, SQLite, HTML, CSS, JavaScript, Django Templates, pypdf, automated testing
Repository:
View Source Code
Demo:
Watch Project Demo
A full-stack customer-support system combining a customer-facing chatbot with a protected business operations dashboard.
The system includes FAQ-based support, lead capture, chat history, unanswered-question management, FAQ administration, analytics, lead status tracking, notes, and CSV exports.
Business users can manage chatbot knowledge and review support activity without editing application code.
Key areas: full-stack development, support operations, admin workflows, lead management, analytics
Tech: Python, Django, SQLite, Django Templates, HTML, CSS, authentication, automated testing
Repository:
View Source Code
Demo:
Watch Project Demo
A Django-based website support and lead-capture system built for a fictional furniture business.
The system answers approved product, pricing, delivery, custom-order, and business-policy questions while detecting commercial intent and capturing structured customer leads for follow-up.
It also demonstrates database-backed lead storage, customer-detail validation, safe fallback behavior, and website-ready support workflows.
Key areas: website support automation, lead capture, intent handling, database workflows
Tech: Python, Django, SQLite, HTML, CSS, JavaScript
Repository:
View Source Code
Demo:
Watch Project Demo
A structured product-data assistant built around an ecommerce product catalogue.
The system answers customer questions about products, categories, sizes, colours, prices, SKUs, stock status, and availability using approved CSV data rather than generating unsupported product information.
Key areas: structured data retrieval, product search, catalogue automation, safe responses
Tech: Python, Streamlit, pandas, CSV, pytest
Repository:
View Source Code
Demo:
Watch Project Demo
A document-based Q&A assistant designed for a fictional ecommerce footwear business.
The system extracts information from business PDF documents, retrieves relevant content, provides source-grounded answers, and avoids answering unsupported questions.
Key areas: document retrieval, PDF processing, source-grounded Q&A, safe fallback
Tech: Python, Streamlit, OpenAI API, PyMuPDF, scikit-learn, pytest
Repository:
View Source Code
Demo:
Watch Project Demo
A controlled FAQ chatbot built for a fictional ecommerce clothing business.
The system answers common customer-support questions from an approved FAQ knowledge base, logs customer questions, and uses safe fallback behavior when information is unavailable.
Key areas: FAQ automation, controlled responses, customer support, logging, testing
Tech: Python, Streamlit, JSON, CSV, pytest
Repository:
View Source Code
Demo:
Watch Project Demo
My next portfolio phase is focused on manufacturing operations and workflow automation.
The objective is to build increasingly complete systems that demonstrate how automation can improve information flow and operational control across manufacturing functions such as:
- Procurement and purchasing
- Inventory and stores
- Production planning and control
- Work-in-progress tracking
- Shop-floor workflows
- Quality control and rework
- Maintenance operations
- Packing and dispatch
- Logistics coordination
- Operational reporting
- Management dashboards
- Cross-department workflow integration
These systems will use n8n, Python, Django, APIs, databases, and AI-enabled components where appropriate, with emphasis on solving operational problems rather than adding AI unnecessarily.
The long-term objective is to develop the capability to design and implement integrated manufacturing workflow automation systems spanning multiple departments and business processes.
- Python
- Django
- Streamlit
- HTML
- CSS
- JavaScript
- n8n
- REST APIs
- Webhooks
- Business process automation
- Workflow orchestration
- System integration
- SQLite
- Structured business data
- CSV / spreadsheet workflows
- pandas
- PDF processing
- Retrieval systems
- OpenAI API
- AI-assisted workflow components
- Document processing
- Classification
- Summarization
- Retrieval-augmented workflows
- Safe fallback and controlled-response design
- Git
- GitHub
- Automated testing
- Requirements analysis
- Workflow design
- Data validation
- Environment configuration
- Documentation
- Client handover preparation
I am building toward specialization in:
AI & Workflow Automation for Manufacturing Operations
My focus is not on automating one isolated task. The goal is to understand how operational processes connect across a manufacturing business and design systems that improve:
- process efficiency
- data accuracy
- operational visibility
- workflow coordination
- exception handling
- management decision support
- integration between business functions
All portfolio businesses, customer records, operational records, orders, products, conversations, and other business data used in these projects are fictional or synthetic unless explicitly stated otherwise.
The projects are designed to demonstrate practical software-development and workflow-automation capabilities without misrepresenting mock systems as production client deployments.
LinkedIn:
https://www.linkedin.com/in/sobangrewal/
GitHub:
https://github.com/sobangrewal479
Email:
sobangrewal.dev@gmail.com