MemGen is a platform that intelligently compiles customized content, converging career chronicles and capacities, allowing users to generate custom cover letters and resumes using large language models and embeddings.
MemGen enables users to create custom and tailored cover letters by searching through their past experiences using large language models and embeddings.
Writing unique and customized cover letters can be a challenge, especially for programmers. The process can be tedious and, although sometimes optional, not writing a cover letter can lead to feelings of guilt. MemGen was created with the goal of providing a platform that allows users to generate custom cover letters using large language models and embeddings, streamlining the process and making it more efficient.
- Use vector databases to store past experiences and search through them using NLP
- Allow users to create custom cover letters using large language models and stored data
- Provide a user-friendly interface for uploading documents and generating cover letters
- Frontend: React, NextJS, Tailwind CSS, Axios, Material UI, Auth0
- Backend: ExpressJS, OpenAI, Firebase Admin, Auth0, Cohere, Google Cloud
- Database: Milvus, Firebase, Zillis
- Integrating multiple technologies and databases to create a seamless experience
- Learning and implementing a vector database
- Ensuring the security of the platform
- Learning and implementing brand new APIs
Through the development of MemGen, we gained experience in working with various technologies and databases. Some of the key learnings include:
- Using vector databases to store embeddings and provide faster search results
- Integrating multiple technologies to create a seamless experience
- Ensuring the security of the platform
- Integrate Stripe API for billing
- Develop a custom vector database for better performance and scalability
- Improve the user interface for a better user experience
- Implement additional features such as job posting and job search functionality