An AI-powered Forward Deployed Engineer for small businesses — built for the Agents League Hackathon 2026.
Small businesses have data everywhere:
- Shopify sales numbers
- Zendesk support tickets
- Inventory spreadsheets
But nobody knows:
- Why did sales drop this week?
- Which products need reordering?
- Which customers are unhappy?
They can't afford a data analyst. They don't have time to connect the dots manually.
OpsPilot AI acts as a Forward Deployed AI Engineer for small businesses.
Ask it a business question in plain English. It automatically pulls from all your data sources, reasons across them, and gives you specific recommendations with numbers — in seconds.
You ask:
Why did sales drop this week?
OpsPilot answers:
Sales dropped 18.2% because the Lavender Diffuser (SKU-1142) hit zero stock on June 8, causing 23 unfulfilled orders. Zendesk out-of-stock complaints spiked from 8 to 29 tickets (+262%). Recommended: reorder 58 units immediately, notify waitlisted customers, and pause low-ROAS ads.
User Question
↓
FastAPI Backend
↓
Azure AI Foundry Agent (GPT-4.1-mini)
↓
Automatic Tool Calling
├── 🛍️ Shopify Sales Tool
├── 🎫 Zendesk Tickets Tool
└── 📦 Inventory Status Tool
↓
Reasoned Analysis + Recommended Actions
opspilot-ai/
├── data/
│ ├── shopify.json # Mock Shopify sales data
│ ├── zendesk.json # Mock Zendesk support tickets
│ └── inventory.json # Mock inventory levels
├── tools/
│ ├── shopify.py # Shopify data tool
│ ├── zendesk.py # Zendesk data tool
│ └── inventory.py # Inventory data tool
├── frontend/
│ └── index.html # Chat UI
├── main.py # FastAPI backend
├── agent.py # Azure AI Foundry agent (CLI version)
├── demo.mp4 # Demo video
└── README.md
Built on Foundry IQ using:
- Azure AI Foundry Agent Service
- GPT-4.1-mini with multi-tool calling
- Automatic agent orchestration across data sources
- Tool calling with real-time data retrieval
| Layer | Technology |
|---|---|
| Agent | Azure AI Foundry (Foundry IQ) |
| Model | GPT-4.1-mini |
| Backend | Python + FastAPI |
| Frontend | HTML / CSS / JS |
| Auth | Azure DefaultAzureCredential |
- Clone the repo:
git clone https://github.com/chayachandana/opspilot-ai.git
cd opspilot-ai- Install dependencies:
pip install azure-ai-agents azure-ai-projects azure-identity fastapi uvicorn python-dotenv- Create a
.envfile:
PROJECT_ENDPOINT=your_azure_foundry_endpoint
MODEL_DEPLOYMENT_NAME=gpt-4.1-mini
- Run the app:
uvicorn main:app --reload- Open your browser at
http://localhost:8000
- "Why did sales drop this week?"
- "Which products need to be reordered?"
- "What are customers complaining about?"
Most AI tools give generic answers. OpsPilot:
- Reasons across multiple data sources in a single query
- Cites its sources — every conclusion backed by data
- Gives specific numbers — not "sales dropped" but "sales dropped 18.2%"
- Recommends exact actions — not "reorder stock" but "reorder 58 units of SKU-1142"
Chaya Chandana GitHub · LinkedIn
Built for Agents League Hackathon 2026 — Enterprise Agents Track