This App Wanted $700 So I Built it Myself with Python

Use Promo Code "PyCharm4Lewis" for a Free 3-Month Personal Subscription for PyCharm, the IDE designed with data and ML professionals in mind: https://jb.gg/PyCharm_Lewis
https://jb.gg/Check_out_PyCharm

When the best software costs $700, you rebuild it with Python instead. 🐍

In this video, I create a voice dictation app that uses state of the art models locally on your machine and processes it to give the most accurate results. We also add other features on top of it just for fun :)

Full scope:
Build a speech-to-text system using Insanely Fast Whisper 🗣️
Create custom keyboard shortcuts⌨️
Process it with AI to get smart formatting🤖
Implement screenshot based text-recognition for better accuracy.📷

Let me know if you want me to open source this project :) I already opened up a pull request on "whisper-writer" which by the end of this video, I started to fork from.

Whisper-writer: https://github.com/savbell/whisper-wr...

LINKS
---
MY 12K+ DISCORD 💬
  / discord  

CONNECT WITH ME ON SOCIAL
📸 Instagram:
  / lewismenelaws  

🎚TikTok:
  / lewismenelaws  

🐣 Twitter:
  / lewismenelaws  

My gear 💻
https://liinks.co/lewismenelaws
-----

TIMESTAMPS
0:00 The $700 Software
0:20 Let's Build it with Python
0:27 Voice Recognition
0:57 Configuring Whisper
1:25 Try PyCharm Today
2:06 Inserting Keyboard Shortcuts
2:33 Demo of First Version
2:54 Adding Voice Post-Processing
3:34 What I Tried First...
3:54 Large Language Model to Post Process
4:03 Demo of 2nd Version
5:36 Providing Context Through Screenshot captures
5:55 Demo of the OCR feature
6:58 Contributing my findings
7:12 Thank you!