Ango AI FAQs

Frequently Asked Questions

[dipl_faq_page_schema closed_toggle_icon=”%%43%%” closed_icon_color=”#cccccc” closed_use_icon_font_size=”on” closed_icon_font_size=”25px” open_icon_color=”#ffffff” open_use_icon_font_size=”on” open_icon_font_size=”25px” closed_toggle_bg_color=”#ffffff” open_toggle_bg_color=”#ffffff” _builder_version=”4.6.1″ _module_preset=”default” faq_heading_text_align=”left” faq_question_closed_font=”Poppins||||||||” faq_question_closed_text_align=”left” faq_answer_text_font=”Poppins|300|||||||” faq_answer_text_font_size=”17px” faq_answer_text_line_height=”1.6em” width=”80%” module_alignment=”center”][dipl_faq_page_schema_item faq_question=”What does Ango AI do?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

Ango AI labels data.

Our customers send us raw data, for example in the form of image, audio, or video, and we label it according to the specifications set by them. The labeling is done by our team of human labelers, whom we call Ango Masters, on our own platform, called Ango Hub. When the labeling is done, we send our customers their labels in the file format they wish.

[/dipl_faq_page_schema_item][dipl_faq_page_schema_item faq_question=”What’s the process of working with Ango AI, from start to finish?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

First, we ask you a series of questions aimed at understanding how you want your labels to be done. This includes label types, data quantity, quality, and other project properties. Then, we ask that you send us a small data sample, which we will label in order to determine the best pricing and deadline for your project, as well as allowing us to make internal preparations for your project.

Once you have started sending us the rest of your data, we dedicate a project manager to your task, who will distribute your data to our team of labelers. We provide you with a username and password, with which you can log in to Ango Hub to check the labeling progress yourself. If you so request, we may create a Slack channel where you will be able to directly contact your project manager for any questions you might have. At any point, you may request the labels that have been completed so far. By the end, you will receive your labels in the format that you request.

[/dipl_faq_page_schema_item][dipl_faq_page_schema_item faq_question=”How does Ango AI ensure their labels are of high quality?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

Firstly, we assign a project manager to each of your labeling tasks, whose responsibility is to ensure that your labels are of the quality agreed upon. If for any reason the labels do not reach the quality standard agreed upon, our project manager will ensure that additional re-labeling is done at no extra cost to you. We employ full transparency during the labeling process, ensuring that you are able to see the labeling progress at any time as well as a number of metrics regarding the quality of the work performed by our labelers, such as their consensus score, time spent on each individual label, and more.

We then provide a consensus system, where more than one person labels the same data point.

For example, assume you tell us that 10 labelers should label the same data point, in this case, pictures of dresses. We then show the same picture to 10 people. If nine of them say the dress is blue and one says it’s gold, then we will accept “blue” as the correct and final answer, with a 90% consensus score. Our platform supports any number of labelers working on the same data point.

[/dipl_faq_page_schema_item][dipl_faq_page_schema_item faq_question=”How does Ango AI ensure my data is stored securely?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

Our customers range from small startups all the way to national security companies. Because of this, we offer different security packages according to the customer’s needs. We go from making use of our remote labeling workforce connecting through HTTPS to our server, all the way to installing our software on your own data center and having our labelers connect securely to your own VPN. If necessary, we can even simply provide the software for a flat fee and you can use it with your own labelers. For more information, we invite you to contact us to discuss your particular needs.

[/dipl_faq_page_schema_item][dipl_faq_page_schema_item faq_question=”Who are some of Ango AI’s typical customers?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

Any company that makes use of machine learning models, or that plans to do so in the future, is a potential Ango AI customer. We have had customers from the e-commerce, customer sentiment analysis, insurance, security, traffic management, and workplace productivity industries, as well as academic projects.

[/dipl_faq_page_schema_item][dipl_faq_page_schema_item faq_question=”Does Ango AI provide services other than data labeling?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

Ango AI prefers to only focus on the process of data labeling, in order to provide the best possible level of service in that field. Ango AI does not provide the data nor does it set up ML models and/or environments. If a small amount of data cleanup is required, however, we might be able to help.

[/dipl_faq_page_schema_item][dipl_faq_page_schema_item faq_question=”What is Ango Hub?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

Ango Hub is our labeling platform, developed in-house by us. Our project managers and labelers both use Ango Hub to provide you with the final product: the labels. Ango Hub supports all kinds of labeling workflows, ranging from simple image classification all the way to named entity recognition with nested answers, speech-to-text, and much more. We would be thrilled to give you a free, no-hassle demo of Ango Hub whenever you’re available.

[/dipl_faq_page_schema_item][dipl_faq_page_schema_item faq_question=”Who are the Ango Masters?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

Ango Masters are our labeler workforce. At Ango AI, every single data point you send us is checked by a human being. While we do make use of AI assistance in order to boost our Masters’ performance, ultimately the label is drawn by the Master. Our Masters are highly selected and trained in their tasks. We assign the most appropriate Masters to your labeling task. For example, there are Masters who are specialized in drawing bounding boxes while others do better in text tasks. By assigning the right Master to the right job, we ensure the highest quality in the shortest amount of time. Most of our Masters are either university students or fresh graduates. For more on our workforce, click here.

[/dipl_faq_page_schema_item][dipl_faq_page_schema_item faq_question=”How can I send you my raw data?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

There is no limit on how you can send us your raw data. We accept anything ranging from a drag and drop form on our own platform, to working directly with your data on the cloud (AWS, Google Cloud Platform, etc.), on your own server through FTP, all the way to having an encrypted HDD shipped to us for maximum security. Feel free to contact us for more information regarding your specific needs.

[/dipl_faq_page_schema_item][dipl_faq_page_schema_item faq_question=”Is there a minimum or a maximum amount of data I can send for a project with Ango AI?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

No. One of Ango AI’s value propositions is that we take on projects of any size, from the smallest to the biggest. We have dealt with projects ranging from 20’000 to millions of data points.

[/dipl_faq_page_schema_item][dipl_faq_page_schema_item faq_question=”What is your pricing model?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

Ango AI prices by the data point. A data point can be one image, 50 words, 1 frame, and others. At the start of the project, we assign a fixed labeling price to your data points. This way, you will know exactly how much the project is going to be priced by the end. 

We do not have a set pricing table because labeling conditions vary for each project.

[/dipl_faq_page_schema_item][dipl_faq_page_schema_item faq_question=”Does Ango AI label with humans or with an AI?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

Each and every label is created and checked by a human. While we do utilize AI assistance features in our platform, ultimately it is a human labeler actually submitting the labels.

[/dipl_faq_page_schema_item][dipl_faq_page_schema_item faq_question=”What does Ango mean?” _builder_version=”4.6.1″ _module_preset=”default” question_heading_level=”h3″]

An ango, or kessei, is a Japanese term for a three-month period of intense training for students of Zen, lasting anywhere from 90 to 100 days. During an ango, a student meditates, studies, and works.

We do not have a set pricing table because labeling conditions vary for each project.