mySupplyChain.ai - AI Business Analysis Platform

Project Overview

mySupplyChain.ai is an AI-powered business analysis platform that transforms how software projects gather and manage requirements. Instead of traditional requirements documents, stakeholders interact with an intelligent Business Analyst Agent that guides them through project definition, asks clarifying questions, and automatically generates structured requirements.

The platform supports multiple AI providers (Claude, Grok, OpenAI), allowing organizations to choose their preferred AI backend. It features a two-step project creation flow: first analyzing the project scope to identify requirement categories, then generating detailed requirements with configurable depth. Real-time progress updates via WebSocket keep users informed during AI processing.

Beyond requirements, mySupplyChain.ai generates HTML wireframes with automated screenshots, integrates with Azure DevOps for work item creation, and supports multi-tenant organizations with role-based access control.

Timeline

March 2025 - Present

Role

Lead Developer / Architect

Type

SaaS Web Application

Status

Active Development

Tech Stack

Node.js 22 Express.js 4.18 MongoDB Mongoose 8.0 Socket.IO 4.8 Claude AI Grok AI OpenAI Puppeteer Azure Blob Storage Stripe JWT Auth

Development Metrics

Commits

AI Providers

Multi

Tenant Support

Real-time

WebSocket Updates

Key Features

🤖 AI-Powered Requirements

Submit your project vision, get structured requirements and intelligent follow-up questions. The AI Business Analyst guides stakeholders through comprehensive project definition.

🔄 Multi-Provider AI Support

Switch between Claude, Grok, and OpenAI with automatic cost tracking. Compare outputs across providers and choose the best fit for your organization.

💬 Interactive Q&A System

Iterative stakeholder interviews to clarify and expand requirements. The AI asks intelligent questions to uncover hidden requirements and edge cases.

🎨 Wireframe Generation

AI-generated HTML wireframes with automated Puppeteer screenshots. Visualize requirements before development begins.

🔗 Azure DevOps Integration

Push generated requirements directly as work items to Azure DevOps. Seamless handoff from requirements to development.

👥 Multi-Tenant Organizations

Team management with role-based permissions. Organizations can manage multiple projects with granular access control.

📡 Real-time Progress

WebSocket-based job notifications across page navigations. Stay informed during long-running AI operations with live progress updates.

💳 Subscription Management

Stripe integration for premium features. Flexible subscription tiers with usage-based billing for AI operations.

Innovation: Two-Step Scope Analysis

mySupplyChain.ai introduces a novel analyze → configure → generate flow for requirements creation. When a user describes their project, the AI first analyzes the scope to identify relevant requirement categories (User Management, Reporting, Integrations, etc.). Users then select which categories to include and configure the depth of analysis. Finally, requirements generate per-category with real-time progress tracking. This approach produces more targeted, comprehensive requirements than a single-pass generation.

Technical Architecture

Multi-Provider AI Architecture

The ai-provider.js service abstracts Claude, Grok, and OpenAI behind a unified interface:

Automatic token counting and cost tracking per provider
Runtime switching allows comparing AI outputs
Provider-specific prompt optimization
Fallback support if primary provider is unavailable
Unified response format regardless of provider

Async Job Processing with WebSocket

Long-running AI operations use a sophisticated job manager pattern:

Socket.IO for real-time progress updates to all connected clients
Jobs persist across page navigations via localStorage
Polling fallback for reliability when WebSocket disconnects
Job queue management with priority support
Automatic retry with exponential backoff

Business Analyst Agent

The core ba-agent.js implements the intelligent Business Analyst:

Context-aware conversations that build on previous responses
Structured output parsing for consistent requirement formats
Domain-specific knowledge for software project requirements
Intelligent follow-up question generation
Requirement categorization and prioritization

Project Structure

mySupplyChain_App/
├── public/                    # Frontend SPA
│   ├── index.html            # Main requirements interface
│   ├── project.html          # Project creation wizard
│   ├── dashboard.html        # User dashboard
│   └── js/                   # Client-side modules
├── server/
│   ├── server-enhanced-dev.js # Main Express server
│   ├── ba-agent.js           # Business Analyst AI agent
│   ├── services/             # AI provider, auth, storage
│   │   └── ai-provider.js    # Multi-AI abstraction layer
│   ├── controllers/          # Specialized agents
│   ├── models/               # MongoDB schemas
│   ├── routes/               # API routes
│   └── middleware/           # Auth, permissions, audit
├── deployment/               # Azure DevOps pipelines
├── wireframes/               # Generated wireframe outputs
└── tests/                    # Jest + Playwright tests

AI Provider Comparison

mySupplyChain.ai supports three major AI providers, each with unique strengths:

Claude (Anthropic)

Excellent at structured output and following complex instructions
Strong reasoning for requirement decomposition
Consistent formatting across responses

Grok (xAI)

Fast response times for interactive sessions
Good at creative problem-solving suggestions
Cost-effective for high-volume operations

OpenAI (GPT-4)

Broad knowledge base for diverse project types
Strong at technical architecture suggestions
Familiar to many organizations already using OpenAI

Wireframe Generation

The platform includes automated wireframe generation capabilities:

AI-Generated HTML: Based on requirements, the AI generates HTML wireframe mockups
Puppeteer Screenshots: Automated screenshot capture of generated wireframes
Azure Blob Storage: Wireframes stored in Azure for easy sharing and versioning
Responsive Previews: Desktop and mobile viewport screenshots
Iteration Support: Refine wireframes through conversation with the AI

# Wireframe Generation Flow
User describes UI requirements
AI generates HTML/CSS wireframe code
Puppeteer renders the HTML in headless Chrome
Screenshots captured at multiple viewport sizes
Images uploaded to Azure Blob Storage
Links returned to user for review
User provides feedback, AI iterates

Azure DevOps Integration

Seamless integration with Azure DevOps for work item creation:

Work Item Creation: Push requirements as User Stories, Features, or Epics
Hierarchy Support: Maintain parent-child relationships between items
Field Mapping: Map requirement fields to Azure DevOps fields
Acceptance Criteria: AI-generated acceptance criteria included
Tagging: Automatic tagging based on requirement categories
Bidirectional Sync: Updates in Azure DevOps reflected in the platform

Security & Multi-Tenancy

Authentication

JWT-based authentication with refresh tokens
OAuth integration options
Password hashing with bcrypt
Session management with secure cookies

Multi-Tenant Architecture

Organization-level data isolation
Role-based access control (Admin, Member, Viewer)
Team management within organizations
Project-level permissions
Audit logging for compliance

Deployment

Azure Infrastructure

Hosting: Azure App Service with Node.js 22 runtime
Database: MongoDB with Mongoose ODM
Storage: Azure Blob Storage for wireframes and documents
CI/CD: Azure DevOps pipeline with automated deployment
Environment: Development (dev-aisc.azurewebsites.net)
Payments: Stripe integration for subscription billing

Future Enhancements

Jira Integration: Push requirements to Atlassian Jira
GitHub Issues: Create issues directly from requirements
Test Case Generation: AI-generated test cases from requirements
Requirement Traceability: Track requirements through implementation
Stakeholder Collaboration: Real-time collaborative editing
Export Formats: Word, PDF, Confluence export options
Custom AI Training: Fine-tune on organization-specific terminology
Analytics Dashboard: Insights into requirement patterns and coverage

Key Dependencies

Express.js: REST API server framework
Mongoose: MongoDB ODM for project/user data
Socket.IO: Real-time job progress notifications
Puppeteer: Headless Chrome for wireframe screenshots
@azure/storage-blob: Document and wireframe storage
Stripe: Subscription billing management

← Back to Current Projects