Building the GitHub Agent: Solving Developer Challenges with AI

​Building the GitHub Grading Agent: Automating Project Evaluation with AI

​Join us for a deep dive into how we built an internal GitHub Agent using LangGraph to automate project evaluations submitted on our platform.

​In this session, Manoj Kulkarni, Gen AI Developer at Hidevs, will walk you through how the agent takes in a GitHub repository URL along with predefined evaluation criteria and expected skills — and generates a detailed report assessing how well the submitted project aligns with those benchmarks.

​By leveraging LangGraph for structured orchestration and reasoning, this tool has significantly improved the efficiency and consistency of project grading. It enables us to deliver faster, more transparent feedback while maintaining high standards in evaluation.

​What you’ll learn:

​The motivation behind building the GitHub Agent

​How LangGraph powers the agent’s reasoning workflow

​How the system uses provided evaluation criteria and skills to generate reports

​Real-world impact on grading consistency and feedback quality

​Founded by Deepak Chawla — Perplexity Business Fellow and jury member at Smart India Hackathon — HiDevs has already upskilled 2000+ individuals and collaborated with top institutions like NMIT, MIT Bangalore, and VGU.
Explore more about HiDevs: https://www.hidevs.xyz
Connect with Deepak: https://www.linkedin.com/in/deepak-chawla1307