Let’s now define the application instance. Notion_block_id = os.environ.get("notion_block_id") Notion_api_key = os.environ.get("notion_api_key") from flask import Flask, requestĪllowed_username = os.environ.get("allowed_username")Īssemblyai_api_key = os.environ.get("assemblyai_api_key") Let’s first import some libraries and import the environment variables. We’re now going to see all the steps of the code, but you can find the whole application on my GitHub repo here: We’re going to config the environment variables later. This allows us to share the code without sharing the API keys. As I said before, it’s a Python web application created using Flask framework and deployed on Heroku.Īs always suggested in these situations, we’d better save confidential data like API keys on some environment variables rather than hard-coding them inside the script.
Let’s now write the core of our application. Everything is orchestrated using a Python application created using Flask and hosted on Heroku (or whichever app hosting platform you prefer). This message is then transcribed using AssemblyAI API and the transcription is then appended to a Notion document. This application starts with the user sending a voice message to a Telegram bot. In this article, I’ll show you how to create a simple voice diary using the free version of AssemblyAI API. Their speech-to-text engine is even able to infer punctuation, which has always been a hard task for most of the speech-to-text engines out there. I’ve given it a try and the result really amazed me.
This tech company has created some speech recognition tools that allow you to do several things with a spoken text, like sentiment analysis and transcription. There are some applications that allow you to transcribe your voice into text, but a really useful tool I found is AssemblyAI. There are several ways to do such a task. It can be really useful if you are busy and don’t have time to write text on your phone and want to take some notes on the fly just recording your voice. The tool I’m about to create is a voice diary, that is a document with notes that have been transcribed starting from voice notes. I’ve decided to try it in order to create a small voice diary using Telegram, Python and Notion.
Recently, I stumbled upon a tool called AssemblyAI, which implements some APIs for speech recognition and analysis.
0 kommentar(er)
