.png)
5 minute read
Meet AutoLettria: Text Classification Made Easy
What if you could train your own artificial intelligence system, on your own data, without having any data science expertise? On top of that, what if you were able to even outperform ChatGPT on your specific use case?
Big changes are coming in how we train text classifiers, and Lettria is leading the way. Today, we want to introduce you to AutoLettria, our new AutoML platform for text – it's simple, fast, and easy to use.
Today's Challenges with Text Classification Training
Right now, training a text classifier is a hard task.
It takes a lot of time, needs a deep understanding of machine learning, and it’s frustrating jumping between different tools for data import, data cleaning, annotation and labelling, and finally model training and testing.
For many companies, these challenges feel like walls they can't get over — and so it’s easy to think of text processing and AI as something that you can’t have unless you invest in data science experts that can lead the way. Not to mention waiting for months if not years for a solution to be developed.
You have to try different tools and techniques, and keep changing your approach until you get it right. It's like running a race where the finish line keeps moving. You invest a lot of time and effort, but you're never sure you'll succeed.
Sure, you could train a deterministic model much more easily, but that’s not likely to give you all the results you’re looking for.
It's clear: we need a new way to train text classifiers.
A New Way to Train Text Classifiers
This is where AutoLettria comes in. Our goal is to make training text classifiers simple and efficient. AutoLettria is our answer to this difficult problem.
With AutoLettria, you can take annotated data in your Lettria project, and automatically train a machine learning model with that data, without writing a single line of code on your own.
Whether you're a skilled data scientist or a business analyst who knows a bit about data, you can use it to build your own text classification model, and deploy it in your application easily. It's a tool that makes machine learning easy for everyone to use.
Here’s how it works —
AutoLettria is automatically an option in any project that has the Text Classification module enabled.
- Upload your data into your Lettria project.
- Set up your annotation labels and launch an annotation campaign. We have an article with advice on label creation here.
- Once your data is annotated, you can launch AutoLettria to build a machine learning model that will classify your data by learning from your annotations.
- You can launch multiple model trainings — test different models that you’ve trained, and evaluate the most accurate one.
- Deploy your solution into production using the API key for the project.
Before Lettria, training your own text classifier would take weeks of setting up, using different tools, and lots of trial and error. But now, you can start your model training in the evening, and by breakfast the next day, you’ll have a first draft that you can already fine-tune and deploy.
This way, you can deploy quickly, iterate quickly, and build a higher quality solution in lesser time.
No more long waiting times and switching between different tools. The days of juggling different systems and spending a lot of time on trial and error are over.
AutoLettria: The Future of Text Classifier Training
Whether you’re training a classifier to automatically detect information in call transcripts to add into specific fields in your CRM, or you’re trying to analyze online reviews for your products, AutoLettria will make your work much faster and allow you to have more accurate results.
AutoLettria is currently still in internal beta, but it’ll be out for everyone to use very soon. Follow us on LinkedIn to stay updated when we launch.
If you have any questions, feel free to reach out to us at hello@lettria.com and we’ll be more than happy to help you out.
