Cordial

Project Description

Cordial is an AI product manager that prioritizes the corporate team structure by leveraging long-lived contextual windows and multi-agent autonomy to proactively manage teams and workspaces.

As a web application, teams can integrate their project, files, and third-party services, to store across widely-regulated cloud services including Google Cloud Provider, MongoDB Atlas, and Weaviate. The core idea is an AI agent, powered by Google Gemini, that actively listens to the conversation. Based on context of the team and their progress, Cordial automatically creates, updates, and assigns tasks/issues directly within GitHub repositories and Jira projects. This lets team members manage workflow using natural language in the chat, reducing friction and keeping everyone synced.

Problem Solved: Reduces the manual effort and potential for errors when translating team conversations into actionable tasks in developer platforms like GitHub and Jira. Makes project context (files, history) readily available to the AI for more informed interactions.

Tracks Entered:

Cold Hard Cache Google Gemini: Chase the Future! MLH: Best Use of MongoDB Atlas MLH: Best Use of Gemini API

Code Repository

Link: https://github.com/kylebtran/cordial

Technologies Used

Backend: Python AI: Google Gemini API (using Gemini-2.5-Pro-latest for chat and backend analysis) Database: MongoDB Atlas (User credentials, projects, relationships, etc.) Integrations: GitHub API (via PyGithub library) Jira Cloud API (via Python jira library)

Frontend: [Specify Frontend Technology - e.g., React, Vue, Flask/Jinja, etc.] Configuration: Python-dotenv

TZwyxoIx-IiMGxMyJ0Vsmg TlgYnesMYWkG1dOkyt5uRQ fewGpoIgJEydIoj_r6y5bg

kylebt1@uci.edu vishokl@uci.edu singhaditya@csu.fullerton.edu elchin4@uci.edu