BIDARA : Bio-Inspired Design and Research Assistant

BIDARA is a GPT-4 chatbot that was instructed to help scientists and engineers understand, learn from, and emulate the strategies used by living things to create sustainable designs and technologies.

🚀 Try it online! or Use BIDARA in ChatGPT

BIDARA can guide users through the Biomimicry Institute’s Design Process, a step-by-step method to propose biomimetic solutions to challenges. This process includes defining the problem, biologizing the challenge, discovering natural models, abstracting design strategies, and emulating nature's lessons.

🔥 Features enabled

     ✅ Multiple chats
     ✅ Code Interpreter more info, filetypes supported
     ✅ Knowledge Retrieval more info, filetypes supported
     ✅ Function Calling
         ☑️ Retrieve academic literature with Semantic Scholar
         ☑️ Generate images with DALL-E
         ☑️ Analyze images with GPT-4V(ision)
         ☑️ Detect pABC patterns in images with GPT-4V(ision)

💻 For developers

bidara-deep-chat uses Svelte and the deep-chat web component to connect to BIDARA over the OpenAI Assistants API. Template based on https://github.com/sveltejs/template

Run locally

npm install
npm run dev

Known issues

• deep-chat speechToText: submit command word is sent in message on safari and chrome on iOS.

• deep-chat textToSpeech: doesn't read messages aloud on safari or chrome on iOS.

Nice to haves

• save chat logs.

• ability rate responses and add feedback.

• ability to send ratings, feedback and chat log to us.

• don't do TtS unless StT has been used.

• Proxy requests to OpenAI through an authenticated API. Users can request access. Ability to generate api keys once authenticated. Authorized API keys required to communicate with API.

• Show the quote from the file used to generate the response when BIDARA uses knowledge retrieval. https://platform.openai.com/docs/assistants/how-it-works/message-annotations

• GPT-4 vision support so it can 'see' images, including ones uploaded by users.

• Functions:

• patent search - https://developer.uspto.gov/api-catalog/bulk-search-and-download, https://patentsview.org/apis/api-faqs, https://www.npmjs.com/package/uspto-patents-view-api

• get pdf of paper from semantic scholar link and upload to assistant.thread.messages.files for retrieval. as a temp workaround can use openAccessPdf links to download pdfs on client, and then upload them on client directly to openai assistant.

• get all messages using openai API. Then BIDARA can use the list of messages to summarize the conversation, or save the conversation history to PDF.

• get all code interpreter code. useful to check its work..