I’ve been pretty inspired by the likes of http://bolt.new and http://lovable.dev for building applications.

So much so, that we’ve started building an interface on our startup’s data extraction capabilities into a rapid data app.

The idea is to build a fast, flexible and intuitive way to go from having a document, to a flexible data capture pipeline.

The demo page is at https://go.sea.dev/ with a video and demo app.

Three key features

1. Generate a schema from a document or description (the target of the data extraction)

2. Logic and “expert insights” get added to field level descriptions/prompts

3. Ability to share with externals to get data (calendly for data requests!)

The primary feature currently lacking is the ability to iterate on the schema using the chat, it is extraction/RAG/FAQ only.

Some other WIP features:

- Improvements on LLM powered OCR

- Email or WhatsApp data capture

- Embeddable data collection for SaaS

Would love to get your feedback and comments!