Skip to content

Latest commit

 

History

History
67 lines (47 loc) · 3.02 KB

README.md

File metadata and controls

67 lines (47 loc) · 3.02 KB

Lookout

Lookout is an open source AI-powered project management platform that streamlines project planning, execution, and monitoring. Teams can offload cognitive capacity to the AI-powered assistant for a range of tasks, including writing story descriptions, querying requirements, and keeping stakeholders informed.

image

Lookout was our capstone project for CS 467 at Oregon State University. This repo highlights our collaborative efforts towards building a full-stack web application and implementing an AI copilot. If you're interested, you can check out how we planned our sprints and groomed our stories here.

Quickstart

Requires Node.js 20 or higher. Run the following commands:

git clone https://github.com/lookout-team/lookout.git
cd lookout
npm install
npx prisma generate

Then, configure your .env file. It should be located in the root directory, /lookout/.env.

DATABASE_URL="file:./dev.db"
OPENAI_API_KEY="YOUR_API_KEY_HERE"
AUTH_SECRET="UKZDaFpcU2mDTHnjsfbDtGFE3V1LP/YQaB2MQtuorts="

To run the app on your local machine, use:

npm run dev

And navigate to https://localhost:3000/home. Have fun :)

Tech Stack

  1. Framework: Next.js
  2. Language: TypeScript
  3. UI: NextUI, Tailwind CSS
  4. Database ORM: Prisma
  5. LLM: OpenAI GPT-4o
  6. Authentication: Auth.js
  7. Testing: Jest

Next Steps?

  1. Enhance UX of components rendered in conversations with the Assistant. With a tight deadline of only 3 development sprints, we had to opt for basic key-value rendering in the end. However, components such as the sprint board or task card could have easily been reused, providing a much richer user experience! Additionally, we might allow the user to edit the data within the component and compose the message to the Assistant ourselves during the iteration process.
  2. Integrate AI-assisted completions in sprint planning and task drafting. It could save the user a lot of time to have a first draft ready for you instead of starting from scratch.
  3. Implement LLMs from other AI Assistant providers. Gives users the freedom to choose which provider they'd like to pay for.
  4. Integrations. Allow users to sign in and import projects from GitHub, Bitbucket, GitLab, etc.
  5. Productionize and deploy!

Screenshots

Home

image

Project Sprints

image image

Task Details

image

AI Assistant

image