Zero-Touch-helpdesk

Zero-touch IT helpdesk bot for Slack Automates 40% of repetitive tickets (password resets, Wi-Fi issues)

Comment 2

🌟 Inspiration

As a passionate physics student with a love for technology, I wanted to solve a real-world problem using AI. IT helpdesks often deal with repetitive requests like password resets or connectivity issues , Proactive system alerts (storage, VPN, updates) Salesforce ticket creation with every interaction. What if we could automate those using LLMs? That idea sparked the creation of Zero-Touch Helpdesk — a Slack bot that acts like your mini virtual IT assistant.

🤖 What it does

Zero-Touch Helpdesk is a smart Slack bot that:

  • Allows users to request help via /helpdesk or /ticket
  • Classifies user issues using OpenAI GPT (e.g., "My Outlook won't open")
  • Returns a summary, category, and actionable suggestion in real-time
  • Offers one-click actions like resetting passwords or getting WiFi help

It reduces the human burden in support operations while offering fast, friendly assistance.

🛠️ How i built it

  • Language: Python
  • Framework: Slack Bolt with Socket Mode for real-time interactivity
  • AI Model: OpenAI's gpt-3.5-turbo via openai-python SDK
  • Environment Handling: dotenv
  • Monitoring: Prometheus Client for basic metrics

Code was structured around Slack commands, interactive message blocks, and LLM-powered ticket classification.

🚧 Challenges i ran into

  • Adapting to the latest OpenAI Python SDK (1.x), which had breaking changes
  • Ensuring the LLM returns valid JSON consistently
  • Slack's interactive payload format and token setup was tricky at first
  • Handling the project solo within a short deadline while balancing video, Git, and documentation

🏆 Accomplishments that i'am proud of

  • Built a working Slack bot that integrates with GPT to generate structured output
  • Designed a real use-case application, not just a demo
  • Learned how to use Socket Mode, interactive Slack features, and LLM prompts effectively

📚 What i learned

  • How to build production-ready Slack bots in Python
  • How to use OpenAI’s latest SDK and prompt engineering for structured outputs
  • Better debugging skills and error handling with AI integration

🚀 What's next for Zero-Touch-helpdesk

  • Fully polish the OpenAI integration and make it more reliable
  • Add integration with real IT ticketing systems (e.g., Jira, ServiceNow)
  • Introduce a web dashboard or admin panel
  • Improve prompt quality and add multi-language support
  • Salesforce ticket creation

⚠️ Note: Due to time limits, the OpenAI part may not be working 100% right now, but it's functional in structure and being actively improved.

Built With

Share this project:

Updates

posted an update —

My Contribution I designed and implemented the entire Zero-Touch Helpdesk system using Slack as the interface and integrated it with OpenAI's LLM (Large Language Model) to automate IT support ticket handling. I built a custom Slack bot using Python, the Slack Bolt SDK, and OpenAI's GPT-3.5-Turbo model to classify issues, suggest categories, and provide immediate responses like password resets or WiFi troubleshooting steps.

I also handled:

Environment configuration with .env files

Secure API integration with OpenAI

Error handling and logging

Natural language classification using AI

Metrics exposure with Prometheus

Demo preparation and documentation

Although my LLM integration is still a work in progress, I’ve laid the full foundation for a scalable, intelligent helpdesk system — and I’m committed to improving it even further.

Log in or sign up for Devpost to join the conversation.