2025 SCaLE 22x - Demystifying Building Natural Language Processing ML Models and How to Leverage Them By Example

Welcome to landing page for workshop Demystifying Building Natural Language Processing ML Models and How to Leverage Them By Example at 2025 SCaLE 22x.

What to Expect

This workshop intends to provide an introduction to running a representative NLP application:

• building a simple NLP model for the first time • looking at a more complex NLP model • building a real-world Voice Assistant Application using multiple NLP models

All of this can be done without requiring GPU resources or lengthy training times. This intends to focus on one of the many ways to attack building NLP models and explore how to do it in practice.

Prerequisites

There are 2 options available to participate in this workshop:

Option 1: The Easy Way to Participate

If you are just looking to dip your toe into learning about Natural Language Processing (NLP) models, we will offer several Google Colab Notebooks OR Kaggle Notebooks where you will be able to participate in the workshop without requiring any prerequisites software or configuration. You just need a laptop and a browser.

However, by choosing this option, you will miss out on the full experience of this workshop (the last section, building a Voice Assistant in particular). There will be a Google Colab Notebook that will get you most of the way there for this section but not give the whole experience.

If you choose this option, click COLAB_TOC.md.

Option 2: A Full Production-like Experience

If you opt for the full experience, you need to have a Linux/MacOS based operating system. Should have already installed the software components in the pre-workshop instructions. You minimally need a functioning Python 3.10+ environment and also need to have portaudio installed.

If you haven't setup anything yet and still want to go with this option, the link below will have the all required software you need to install.

If you choose this option, click LAPTOP_TOC.md.

SaaS Platform Accounts Need for Either Option

For the last third of the lab, we are going to make use of the following SaaS platforms:

• Deepgram - This is to help facilitate building an Voice-Activated AI Assistant/Agent.
• OpenAI - This is to provide an LLM in the examples.

Deepgram Account

You need a Deepgram account in order to do the Speech-to-Text and Text-to-Speech. The account is free and you get $200 in credits! Sign up at Deepgram.com

OpenAI Account

You need a OpenAI API Key to use an LLM to provide responses in the Voice Assistant. Sign up at OpenAI.