Data Pipeline to Automate PDF File Translations | Google Cloud Gemini

🌐 Automate PDF Translation with Google AI and Gemini!
In this video, I’ll show you how to automate PDF translation using Google Cloud’s AI tools, including Google Gemini. With just a few steps, you can streamline translating large documents, saving time and effort. Whether you're a business professional, student, or developer, this guide will help you harness the power of Google Cloud to handle multilingual content with ease.
Source Code
www.techtraptu...
#translation #gemini #ai
Looking to get in touch?
Drop me a line at [e-mail], or schedule a meeting using the provided link topmate.io/vis...
Playlists
Associate Cloud Engineer -Complete Free Course
• Associate Cloud Engine...
Google Cloud Data Engineer Certification Course
• Google Cloud Data Engi...
Google Cloud Platform(GCP) Tutorials
• Playlist
Generative AI
• Generative AI
Getting Started with Duet AI
• Getting started with D...
Google Cloud Projects
• Google Cloud Projects
Python For GCP
• Python for GCP
Terraform Tutorials
• Terraform Associate C...
Linkedin
/ vishal-bulbule
Medium Blog
/ vishalbulbule
Github
Source Code
github.com/vis...
#googlecloud #gcp

[@rishiraj2548]

🙏🏻👍🏻 Nice that you're again with us.

[@techtrapture]

I have never left you😀

[@machu_pichu0232]

Thank you for this video. Is there a way to access the code for the function? The link just has an error.

[@SwapperTheFirst]

Nice to see you again. Are you using Doc AI in this project? Or it is rather to extract text to PDF and send to Flash for translation? Aha, checked drawio - you send text to LLM Model endpoint, not using Doc AI OCR...

[@techtrapture]

Why to complicate if we have easy options now 😀

[@SwapperTheFirst]

@@techtrapture Large LLM is slow _and_ expensive for really big amounts of text. If you have 10 PDFs to translate, it is fine, but if it's 10 large PDFs per minute? And small LLMs will not help you much, and SFT-Lora or Distillation are not feasible for this use case. So it is not a complication, rather what is the _optimal_ tool for your task. The same with VLM and traditional determenistic/DL approach for OCR. You will hit a wall with VLMs very soon on large amounts of OCR. Sure, you can parallelize the OCR, but the price may bring tears to your eyes.

[@klb-og7cp]

@@SwapperTheFirst But it is by far the best quality. DeepL is faster but nobody wants to read that sht lol.