top of page
Here's the feature we are currently building:
Feature: Implement HTML parsing AI model for Layout Engine
ProductMap | Use Case | Difficulty | Benefit | ETA | In Progress | Completed |
|---|---|---|---|---|---|---|
Feature: Implement image parsing LLM for Layout Engine | Being able to understand webpages with image data helps the context engine figure out what to do | High | Very High | August 20, 2024 | X | X |
Feature: Capture and reproduce mouse events | Capture and replaying mouse events should allow us to improve workflows for websites that don't expose HTML/JS directly. Being able to simulate specific actions will help with actions like logging in, SSO, ReCaptcha's, etc. | High | Very High | August 31, 2024 | X | X |
Feature: Add saving to custom screenshot location | Being able to specify where to save screenshots in a workflow helps split where images are being stored | Low | High | July 13, 2024 | X | ✓ |
Feature: Add ability to click on visible text | Core feature | Low | Very High | July 20, 2024 | X | ✓ |
Feature: Add ability to search for text | CTRL + F for interactions is helpful. | Low | High | July 31, 2024 | X | ✓ |
QOL: Create detachable browser | Standalone browser = high scaling. | Medium | Medium | N/A | X | X |
QOL: Create detachable chat | Detaching the chat and letting the browser run in a standalone sandbox should improve scalability once we move to a server-client model. | Medium | Low | N/A | X | X |
Feature: Implement multi-language support | Making it non-english agnostic should be straightforward given how we handle converting user inputs into engine choices | Medium | High | November 16, 2024 | X | X |
Feature: Allow clicking on images based categorization | Being able to click on images once we are able to recognize them should allow for the ability to collect links, download images, etc. | Very High | High | November 30, 2024 | X | X |
Feature: Implement table recognition for CSVs | Once we can recognize images, we can have the context layer work to understand what a "table" is and how to extract that into CSV format. | Medium | Very High | October 31, 2024 | X | X |
Feature: Implement HTML parsing LLM for Layout Engine | Being able to understand webpages with raw HTML helps the context engine figure out what to do | Medium | Very High | September 15, 2024 | ✓ | X |
Feature: Add image recognition | Recognizing images will help with the layout engine and the context engine understanding what's on a given page at any point during a workflow. | Very High | Very High | September 30, 2024 | X | X |
bottom of page