How do you differ from existing solutions on the market?
Machine learning teams in large companies face difficulties in data labeling. They have two options: to make an internal tool that requires to search for hundreds of crowdworkers, or to use existing solutions on the market. Unlike other services, we provide better quality for a lower price, as we have a strong training onboarding and the community of labelers with 24/7 support. Beyond that, we use a consensus algorithm that leaves no room for errors.
What is consensus algorithm?
The consensus algorithm is an internal AI-based mechanism. It ensures quality by assigning the very same task to a number of crowdworkers until the wanted accuracy is reached. Consensus works for any type of tools and any data.
How do you handle data safety?
Only a tiny part of data is shown to a single crowdworker. For example, when labeling ID, we divide it into several fields so that one worker does not see the whole document. Later on, we put the fields together inside the client contour.
What do I need to launch labeling?
The process is simple: you set the task, the crowd labels data with our tools, you check the quality in your personal profile and download a dataset in any convenient format.
What can your crowdworkers annotate?
Our crowd is good for general tasks. You can assign tasks to groups of workers each of which is focused on a specific area. But if labeling requires the knowledge of Chinese or medical education, you may need your in-house team.
Who are your crowdworkers and what are their qualifications?
Handl crowd is a well-trained and qualified workforce to deal with any data preparation. Anyone with a connected device can pass onboarding and start labeling data on Handl. We make sure that only qualified ones get access to labeling. Using quizzes and detailed instructions, we ensure that the job is done to the highest standard.
What languages do your labelers work with?
We work mostly with English and Russian but any common language can be used if needed.
Do you work with videos or just images and texts?
Yes, we do. Usually we cut them in frames and label the resulted images. If you need to classify actions in videos, you can upload them in video clips.
Can I create custom interfaces on Handl?
Yes, Handl interfaces can be mixed together and used for any type of data.
Do you have API channels for ongoing data flows?
Yes, we do provide an API channel for recurring tasks.
Do you have on-premise app?
We are working on it.
How scalable is the product?
If you have an ongoing flow of data for annotation, you can get an API channel and have your data labeled as it comes.
What is the usual pricing?
We have flexible pricing for any type of tasks. Please, check it out here.