CRM & Ticketing System

Reusable, containerized backend for CRM and ticket workflows.

GitHub · Live Demo · Back to Projects

---

• Problem: Demonstrate a reusable CRM/ticket backend with realistic access patterns and clean API boundaries.
• Approach: FastAPI service with RBAC tokens, Pydantic schemas, and containerized deployment.
• Outcome: One backend image reused across UI demos and “agent” clients without API changes.

---

A standalone CRM & Ticketing backend built with FastAPI and packaged as a reusable Docker image.

Designed with clean APIs, UI/API decoupling, and enterprise-style access patterns, making it suitable for both traditional UIs and LLM/agent integrations.

The API runs independently and can be consumed by web interfaces, agents, or other services without modification.

Key Capabilities:

• FastAPI REST API for customers, tickets, and ticket notes.
• Token-based authentication with RBAC (Agent/Admin enforced at the API layer).
• Strict request/response schemas using Pydantic.
• Health/readiness endpoints for uptime checks and deployments.
• Dockerized deployment (reproducible local + portable demo).

---

Architecture

The backend is UI-agnostic and supports multiple clients, including web UIs and AI agents.

            +-------------------+       HTTP        +----------------------+
            |     UI / Agent    |  ─────────────▶   |    CRM API Service   |
            |   (Gradio, LLMs)  |                   |   (FastAPI, Docker)  |
            +-------------------+                   +----------------------+
            

A Gradio UI is layered on top for demo purposes via Hugging Face Spaces. In production-style usage, the backend runs independently and is reused across clients.

---

Demo Deployment

• Docker demo is hosted in Hugging Face Spaces
• Gradio UI communicates with the API over an internal service URL
• Preconfigured tokens simulate enterprise access patterns

---

Engineering Notes