-
semantic-search-openai-pinecone
Semantic search with openai's embeddings stored to pineconedb (vector database)
-
chatgpt-retrieval-plugin
The ChatGPT Retrieval Plugin lets you easily find personal or work documents by asking questions in natural language.
-
SurveyJS
Open-Source JSON Form Builder to Create Dynamic Forms Right in Your App. With SurveyJS form UI libraries, you can build and style forms in a fully-integrated drag & drop form builder, render them in your JS app, and store form submission data in any backend, inc. PHP, ASP.NET Core, and Node.js.
If you want natural language search for your summaries, you could run the results through OpenAI Embeddings and store the results in a semantic vector database like Pinecone. Here is a sample project that does that: https://github.com/mharrvic/semantic-search-openai-pinecone
If the browser plugin for ChatGPT gets better to actually exceed what a web-scraper / summarizer can do, you could use GPT for the summarization part and store the results again in some vector database after creating embeddings. You could then use ChatGPT to "talk to your data", meaning when asked for specific information, it could query the database, retrieve the relevant knowledge and provide a meaningful answer based on your question. This could be done via OpenAIs own retrieval plugin (see https://github.com/openai/chatgpt-retrieval-plugin ) or by creating your own Chat-UI that performs the semantic search, feeds the result into the GPT4 API and returns the answer back to the user. David Shapiro made a nice video with step by step instructions on how to do that: https://www.youtube.com/watch?v=2xNzB7xq8nk&pp=ygUUZ3B0NCBtZW1vcnkgcGluZWNvbmU%3D
Related posts
-
Show HN: Simple email mode for Gmail's 20th anniversary
-
GPT-4 Turbo with Vision is a step backwards for coding
-
Deploy Full-Stack Next.js T3App with Cognito and Prisma using AWS Lambda
-
Building an Admin Console With Minimum Code Using React-Admin, Prisma, and Zenstack
-
Oration (iOS) Turns PDFs into Audiobooks