Inspiration
The idea came while talking about rationality because both members of the team are part of the less wrong community and often try new ways to improve their thinking and seeing the things more cleay
What it does
It is a journaling tool that enables you to improve your decision making and daily life situations
How we built it
Ww built it with next.js for the front end which is deployed in vercel. The backend at the moment are lambda functions in AWS which containe a docker image hosted in ECR. The lambda function communicates with the openAI API and also with the pinecone index where the information from the user is stored. The lambda function also stores some information in a JSON file in an S3 bucket and communicate with the front end via API gateway in AWS. At the moment we also have implemented telegram bot for the mobile interface which communicates with the lambda via the same API gateway
Challenges we ran into
Running langchain in lambda without containers Deciding on the stack Deploying with Vercel from a GitHub organisation
Accomplishments that we're proud of
Having a working product
What we learned
Deploying lambdas with containers About langchain and pinecone About submodules ...
What's next for LogicLog
Many new features like SST, improvements in privacy, ...see GitHub repo for more detailed roadmap :)
Built With
- amazon-web-services
- langchain
- openai
- pinecone
- vercel
Log in or sign up for Devpost to join the conversation.