inmarket-AI-copilot

> An AI-powered, agentic marketing activation platform that discovers real upcoming events via the Ticketmaster Discovery API, scores them against brand/audience fit, and generates complete campaign packages — briefs, copy, and AI-generated poster creatives via Google Gemini — all orchestrated by a LangChain agent through tool calling.

---

• Overview
• System Architecture
• Tech Stack
• Project Structure
• Getting Started
• Prerequisites
• API Keys
• Installation
• Environment Configuration
• Running the Application
• API Reference
• LangChain Agent (LLM Orchestration)
• MCP Server
• Domain Logic
• Intent Normalization
• Event Scoring Engine
• Campaign Brief & Copy Generation
• Image Generation
• Refinement Pipeline
• Frontend (Chat UI)
• Data Models
• Configuration Reference
• Deployment
• Development Notes

---

A marketer types a natural-language prompt like:

> *"Find family-friendly events in Dallas this weekend for a cold beverage brand"*

The system then:

1. Normalizes intent — extracts city, timeframe, brand category, audience, constraints
2. Searches Ticketmaster — queries the Discovery API v2 for real upcoming events by date range
3. Scores & ranks — evaluates each event on 6 dimensions (city fit, audience fit, brand fit, category fit, timing fit, visibility fit) totaling up to 100 points
4. Generates 2 campaign recommendations