Add your PDF, Word, HTML or Text docs, train our AI on them with OpenAI embeddings & vector search and then process results with a GPT3 script. This workflow is perfect for anything NOT in ChatGPT: 250-page compliance PDFs, training manuals, your diary, etc.

Add your PDF, Word, HTML or Text docs, train our AI on them with OpenAI embeddings & vector search and then process results with a GPT3 script. This workflow is perfect for anything NOT in ChatGPT: 250-page compliance PDFs, training manuals, your diary, etc.

Add your PDF, Word, HTML or Text docs, train our AI on them with OpenAI embeddings & vector search and then process results with the LLM script and engine of your choice.

Aimed at Indian Chili farmers, this bot parses 4 documents representing best practices from the Indian Ministry of Culture and Digital Green's work to collect common questions from Indian Chili farmers. We then load these as text embeddings and then run the GPT3 script below to create an answer to the farmer's question, giving citations back to the source documents.

Here's how you can do this for your own account:

1. Request your twitter archive
2. Download archive, unzip, and open Your archive.html in Chrome.
3. Go to the tweets section and scroll to the bottom.
4. Save page as html (Web Page, Complete)
5. Upload it here

Here we use the WHO’s 180 page guide to antenatal care for pregnancies and then can ask any question to receive simple, referenced answers. https://www.who.int/publications-detail-redirect/9789241549912

Here we downloaded a webpage into a PDF and then

1. Parse the PDF into openai embeddings
2. Search the PDF for "What are the most interesting aspects of this character?"
3. Add the results into a GPT3 script that is asked to generate a poll from the most interesting aspects.

Structured SEO metadata is vital to create for SEO (it literally makes your click target bigger in Google's search results e.g. https://developers.google.com/search/case-studies/overview) but it's a pain to create and maintain. This workflow summarizes any webpage with MapReduce and GPT and then creates the schema.org JSON that you can add directly to your page

Add your PDF, Word, HTML or Text docs, train our AI on them with OpenAI embeddings & vector search and then process results with a GPT3 script. This workflow is perfect for anything NOT in ChatGPT: 250-page compliance PDFs, training manuals, your diary, etc.

Here we summarize the downloaded html doc and then create a quiz via GPT

Google Palm vs GPT4 vs LLaMA2: Which can easily create code to call the Stripe API using natural language.

This workflow creates a JSON quiz based on a character.

Here we compare the top 5 ASR models from a set of Telugu samples. Speech output created from https://gooey.ai/bulk/?example_id=nrkx2u17

In this bulk run example, we compare 6 different Gooey.AI speech recognition + translations workflows (https://gooey.ai/speech), each of which uses a different AI speech model:
Whisper Telugu Bhashini
Whisper v2
Google Chirp
Azure
Meta Seemless 4MT
AI4Bharat.org Conformer Hindi

We then evaluate them at https://gooey.ai/eval/ (where you can see what works best....)

Here we feed our workflow an SEC filing and then have it output structured data about the company.