BakeryFlow Agent

Inspiration

Bakery businesses often manage inventory, recipes, suppliers, purchasing, and production across multiple screens and workflows. BakeryFlow was created to simplify these operations.

For the WebMCP Challenge, I am extending BakeryFlow into an agent-native ERP where AI agents can interact directly with real business workflows using WebMCP.

What it does

BakeryFlow Agent allows an AI agent to understand and interact with bakery operations.

Using WebMCP, an agent can:

• Check raw material inventory
• Retrieve product recipes
• Calculate ingredients required for production
• Detect inventory shortages
• Find suppliers
• Prepare or create purchase orders
• Create production runs

For example, a manager can ask:

"Can we produce 1,000 chocolate cookies?"

The agent can inspect the recipe, calculate the required ingredients, compare them against current inventory, identify shortages, and help take the next required action.

How I built it

BakeryFlow uses a React frontend with a Django REST Framework backend and PostgreSQL database.

For this challenge, WebMCP is being integrated into the frontend to expose selected ERP capabilities as structured tools that AI agents can discover and use.

The WebMCP layer reuses BakeryFlow's existing business logic and APIs instead of creating a separate system specifically for AI agents.

WebMCP Challenge Work

BakeryFlow existed before the WebMCP Challenge.

The core ERP functionality, including inventory, recipes, suppliers, purchases, and production management, was already part of BakeryFlow.

For this challenge, I am adding the WebMCP integration and agent-native workflows that allow AI agents to securely interact with those existing capabilities.

Challenges

One of the main challenges is turning complex ERP workflows into clear and safe agent tools while preserving the existing validation, permissions, and business rules.

Another challenge is designing tools that agents can combine into useful multi-step workflows rather than simply exposing individual CRUD operations.

What's next

The goal is to expand BakeryFlow Agent so an AI agent can assist with the complete production lifecycle — from demand planning and inventory analysis to purchasing, production, stock movement, and business insights.

Built With

• django
• javascript
• postgresql
• python
• react
• rest
• vercel
• webmcp