AI-DevOps_Bot

Your AI-Powered Wingman for Smarter, Faster Software Delivery. an AI-powered assistant that enhances your GitLab workflow by automating issue summarization, code review, and test generation.

Comment

Inspiration

Building robust software often involves repetitive tasks like summarising long issue threads, writing test stubs, or reviewing code quality — all of which slow down developer productivity. I wanted to automate these pain points using AI + DevOps — to help teams move faster and smarter.

What it does

AI-DevOps_Bot is an intelligent assistant that streamlines DevOps tasks right from a simple UI. It:

Summarises complex GitLab issues using NLP

Automatically generates test stubs for Python functions

Reviews code using pylint and returns feedback with scores

How we built it

We used:

Streamlit for the frontend app

Transformers (BART) from Hugging Face for summarization

Pylint for static code analysis

Python's ast module for test stub generation

GitLab CI/CD to automate builds

Docker + Cloud Run for deployment on Google Cloud

Challenges we ran into

Truncating long GitLab issue content based on token limits instead of characters

Handling various edge cases in user input

Ensuring smooth CI/CD pipeline integration on GitLab

Setting up a proper Docker-based deployment with minimal latency

Accomplishments that we're proud of

Fully functional working demo with end-to-end automation

Achieved a clean modular architecture

Successfully deployed on Google Cloud

Built a user-friendly experience with Streamlit

Wrote a scoring-based reviewer to mimic real-world code review workflows

What we learned

Writing clean CI/CD pipelines with GitLab YAML

Building and deploying Streamlit apps in production via Docker

Token-based NLP text management for real-world GitLab inputs

Connecting multiple AI modules into a unified app experience

Practical cloud deployment using GCP tools like Cloud Run, Artifact Registry

What's next for AI-DevOps_Bot

Add support for JavaScript, Go, and C++ for review and testing

Integrate GitHub support in addition to GitLab

Add Slack/Discord Bot integration for team notifications

Improve model quality with fine-tuned summarizers

Explore self-hosted options for enterprise users

Built With

  • and-deployed-using-google-cloud-(cloud-run
  • artifact
  • ast
  • cloud-build
  • gitlab-apis
  • hugging-face-transformers
  • langdetect
  • pylint
  • python
  • streamlit
Share this project:

Updates