aiCRM: Next-Generation Agentic CRM

aiCRM is a modern, AI-first Customer Relationship Management system designed to automate sales, client interactions, and task management. Built with a robust Java Spring Boot modular monolith backend and a lightning-fast Next.js React frontend, the system seamlessly integrates Large Language Models (Google Gemini via Spring AI) directly into core business domains.

🚀 Key Features

• Global AI Assistant: Not just a chatbot, but a functional agent. The AI can read your database, create tasks, update deals, and draft emails autonomously using Spring AI's Tool Calling / Model Context Protocol (MCP).
• Omnichannel Integration: Built-in support for receiving and sending messages via Telegram and Email. Incoming messages automatically create new clients and deals without manual entry.
• Kanban Task Management: A drag-and-drop Kanban board to manage daily tasks, integrated directly with clients and deals.
• Deal Pipeline & Auditing: A modern, two-column deal interface (similar to classic CRMs like Pipedrive) that tracks deal stages, multi-currency budgets (USD, EUR, GBP, UAH), and maintains a detailed, scrollable history of events and chat logs.
• Real-time Notifications: WebSockets and long-polling systems deliver instant updates for new messages, AI activities, and systemic alerts straight to your UI.
• Premium Modern UI: Built with Next.js, Tailwind CSS, and shadcn/ui. Features a dark mode, glassmorphism, responsive grids, and optimistic UI updates for zero-latency interactions.

---

🏗 Architecture Overview

The system is split into two major components:

1. Backend (/backend/aicrm):

   • Framework: Java 17, Spring Boot 3.x.
   • Pattern: Domain-Driven Design (DDD) Modular Monolith.
   • Modules: iam (Auth), sales (Clients, Deals, Tasks), communications (Telegram, Email, WebSockets), ai (Spring AI Agents), and shared.
   • Database: PostgreSQL / H2 (via Spring Data JPA).
   • (See backend/aicrm/ARCHITECTURE.md for details).

2. Frontend (/frontend):

   • Framework: Next.js 14+ (App Router), React 18.
   • Language: TypeScript.
   • State Management: React Query (Data Fetching), Zustand (AI Chat State).
   • Styling: Tailwind CSS, Radix UI (shadcn/ui).
   • (See frontend/ARCHITECTURE.md for details).

---

🛠 Getting Started

Prerequisites

• Java 17 or higher
• Maven
• Node.js 18+ and npm
• Google Gemini API Key (for AI features)

1. Starting the Backend

1. Navigate to the backend directory:

   cd backend/aicrm

2. Configure your environment variables (e.g., spring.ai.vertex.ai.gemini.api-key) in application.properties.
3. Run the Spring Boot application:

   mvn spring-boot:run

   The backend runs on http://localhost:8080 by default.

2. Starting the Frontend

1. Navigate to the frontend directory:

   cd frontend

2. Install dependencies:

   npm install

3. Run the Next.js development server:

   npm run dev

   The frontend runs on http://localhost:3000 by default.

---

🤖 AI Integration Details

aiCRM utilizes Spring AI to expose internal CRM domains to the LLM. When you ask the AI to "Create a task to call John tomorrow", the AI leverages the @Tool annotated methods in SalesAITools to execute a direct, secure Java method that updates the database, rather than just returning a text response. The frontend then automatically refetches the data to display the newly created task instantly.

📄 License

This project is proprietary and confidential.