#19 - "We use Sentimental Analysis and AI to get info from tourists in Azerbaijan" 🇦🇿

Azerbaijan, WhisperAI and a new AI course - The AI Coffee Newsletter ☕

Alejandro Pérez Pérez

Apr 06, 2023

🌍 Madrid, Spain

😎 On cloud nine, due to the sunny weather we are having.

🎵 Save your tears - The Weeknd and Ariana Grande.

Hi! Welcome!

This week was quite a nice one, I hope also yours was a good one! 😀

As you might have seen, I created a course for begginers with AI so to improve the efficiency using tools such as Chat GPT and Dalle. Did you like it? Please, leave me feedback in comments or answering the mail so I could improve in the future!

I also changed a bit the design of the newsletter, so now you have the quote of the podcast as the title. Do you like it?

Yesterday I was researching a bit about Whisper AI and how this powerful model for transcripting audio could be implemented to transcript a podcast episode and then use a NLP model to extract the notes of the episode and create a description. At this moment I am only able to do the first part, but as the transcriptions could have like 6000 words for a 30-min conversation, I cannot work with this unless I split it in shorter pieces… and this is not working really well in my code, yet! Any suggestion? (I will leave the code at the end if you want to reuse it)

Today I start a 5-day holiday Easter period, so this week it is going to be mostly a relaxing time to recharge the batteries. I hope you could have some sort of break also 😊

Episode 19 - AI Strategy in Azerbaijan with Khayala

"We use Sentimental Analysis and AI to get info from tourists in Azerbaijan" 🇦🇿

🎧 Want to stay ahead of the curve on AI and digital transformation?

Then be sure to tune in to the latest episode of The AI Coffee Podcast, where host Alejandro Pérez Pérez interviews Khayala Fatullayeva, an expert in AI for public services and a driving force behind Azerbaijan's digital future. With years of marketing experience and multiple master's degrees, Khayala brings a unique perspective to the table as she discusses the challenges and opportunities facing Azerbaijan's digital transformation.

From exploring practical digital solutions that are making a difference in the public sector to expanding our view of AI beyond the traditional powerhouses, this episode is packed with insights and inspiration. So pour yourself a cup of coffee and join us for an engaging and enlightening conversation on The AI Coffee Podcast! ☕

For more information, you can listen to the episode here 👇

(If you do not have Spotify, you can access to the podcast in other platforms from here 😊)

Some useful links ⭐

Linkedin profile: https://www.linkedin.com/in/khayala-fatullayeva/
Awarded Azerbaijan's Electronic Visa System: https://www.evisa.gov.az/en
OECD analysis of Digitalization in Azerbaijan: https://www.oecd-ilibrary.org/sites/6a612a2a-en/index.html?itemId=/content/publication/6a612a2a-en

Bonus: WhisperAI Code 🤖

If you are able to use it in the sense I said before, please let me know!!! 🦄

!pip install git+https://github.com/openai/whisper.git
!pip install pytube
!sudo apt update && sudo apt install ffmpeg

import logging
import pytube
import whisper
import sys
import argparse

parser = argparse.ArgumentParser(description='Transcript a YouTube video using Whisper')

args = {
  "video": input('Pass the YouTube url to transcribe: '),
  "model": 'small'
}

logging.basicConfig(
  level=logging.INFO,
  format="%(asctime)s [%(levelname)s] %(message)s",
  handlers=[
    logging.StreamHandler(sys.stdout)
  ]
)

if not args['video']:
  logging.error("Please pass a YouTube url to transcribe")
  exit()

logging.info("Downloading Whisper model")
model = whisper.load_model(args['model'])

logging.info("Downloading the video from YouTube...")
youtubeVideo = pytube.YouTube(args['video'])

logging.info("Get only the audio from the video")
audio = youtubeVideo.streams.get_audio_only()
audio.download(filename='tmp.mp4')

logging.info("Transcribe the audio")
result = model.transcribe('tmp.mp4')

print(result["text"])

The AI Coffee Newsletter ☕