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.
Our domain expertise lies in building best in class software infrastructure for operational processes. Outsourcing is out of scope for us.
However, our unique position allows us to build relationships with an ever-growing number of Outsourcing firms with deep expertise in operationalising processes of all kinds cost-efficiently. We're happy to provide introductions where appropriate.
Our goal is to empower users to structure their repetitive work and collect data examples as they work, such that training data is generated.
Having a sizeable 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 firstname.lastname@example.org.