check out the general purpose PostgresDB prototype version of Quoordinates I'm working on -- Commonbase!!

• Readwise
• OpenAI Embeddings
• Nomic
• PSQL (PostgreSQL)

1. fetch highlights from readwise
   1. (optional) start at date
2. upsert data in database
3. embed highlights with openai
4. place embeddings into nomic visualization

embed-into-supabase.js is a node script that embeds highlights into supabase nomic-viz.py is a python script that visualizes highlights in nomic readwise-highlights.py is a python script that fetches highlights from readwise

The apps below depend on Quoordinates and display how versatile this technology is:

• Commonplace Bot - A Discord Bot where quotes can be surfaced, discussed, and transformed in an MMO like format.
• Quo Host - A realtime speech to semantic search to "higlight within highlight" surfacer. Sound smarter on podcasts.
• Kindle Streamer - Use OCR on a cron job to take a picture of Kindle, and generate a visual bookmark using DALL-E.

1. Create a .env file with the following variables:

READWISE_ACCESS_TOKEN=
SUPABASE_URL=
SUPABASE_KEY=
OPENAI_API_KEY=
OPENAI_ORG=
DAY_LAST_FETCHED=
LOG_PATH=

2. Install dependencies

poetry init
poetry shell
poetry install

and

npm install

3. Run the script to get readwise highlights

python readwise-highlights.py

4. Run the script to embed highlights

node embed-into-supabase.js

5. Run the script to visualize highlights

python nomic-viz.py

There are other tools in this repo too!

• compile quotes - use k-means to make a cluster of all quotes from a book and give them topic titles
• share-pic - use a quote, dalle, and cf images to save a quote and image in one eye catching piece