Yes and no. We believe any routine manual process can be framed as a data labelling process. This is especially true in the context of Machine Learning based automation.
However, data labelling platforms typically focus on a specific set of process types around image or text annotation. Most manual processes out there are more complex than that, dealing with multiple signals of potentially different data types and decisions that need to be handled together. These are the kinds of problems Human Lambdas can help the most with, due to the unique flexible and customizable nature of the platform.
Yes. Check out the section on Human Lambdas Run, which is our queue execution service. It's like automation, but powered by people.
We don't, but we can help you get there, by giving you the means to accumulate training data so you can feasibly explore automation strategies on your end.
Having a sizable training set is a precondition for automation. We encourage our users to explore ML-based approaches to automate their processes once these reach a more mature, battle-tested phase.
The setup we encourage is a Human in the Loop one, which we have extensively written about:
Human Lambdas is completely free while in Beta.
This will need to change eventually, but we are committed to offering a pricing model that accommodates to each one of our user's unique circumstances while rewarding early adopters.
You can find us at [email protected].