Our second success story is about generating employment for the second broad group of neurodiverse people in the space of AI.
Group 2 includes all those who were diagnosed young with a disability and moved into special schools and classes. They were written off at a young age by the system. Most of them don’t make it past Class X. They can’t hide their differences. Their families in order to keep them engaged, have them attend MS Office and Graphic Designing classes in good faith but most don’t find any paying jobs. They are stuck in an endless loop of training with no jobs in the end, and hence no pay.
Machine Learning is fueled by accurately labeled training data. With the explosion in the practical applications of AI, specifically ML, this group who wants employment can contribute significantly to the creation and annotation of training data for machine learning models. Eventually, we can explore other things in AI Ops like feature engineering.
With the Total Addressable Market being 197 million NDs in India and the global data annotation market was valued at US$ 695.5 million in 2019 and is projected to reach US$ 6.45 billion by 2027 expected to grow at a CAGR of 32.54% from 2020 to 2027. Our plan is to bring these 2 together - the Data Annotation booming market and the untapped resource pool of 197 million and we foresee considerable overlap.
Right now, there is no company in India working at the intersection of neurodiversity and data annotation. We would be creating a new market.
In data annotation, there is a premium on attention to detail, hyperfocus, replicability and volume, and when it comes to image annotation it becomes language agnostic. These are all in sync with the strengths of Group 2.
Data Annotation, particularly, video and image annotation for computer vision, will be the best place to start for this group, in India.
While providing jobs ideal for the target audience, leveraging their strengths - Attention to Detail, High Precision and focus on Accuracy, Hyper-focus, Pattern Recognition, Comfort with Repetitive Tasks and Computer Skills and Honesty and Integrity - will also appeal to corporates looking to outsource their annotation as it would count as strategic impact sourcing that would make sense to business as it would lead to a significant cost saving.
Our plan is to establish Remote-first Data Annotation Units.
We screened a pilot team of 12 NDs on image annotation skills and aptitude, using a tool developed by our ND Tech Team. The team is currently training on Israel-based Data Annotation Platform - Dataloop since 4 Nov’ 21.
The traction received led us to onboard an additional 50 NDs on the platform from 20 Nov’ 21. 42% are either literate with no formal education or have formal education less than X Standard, 21% have completed X, 30% completed XII and 6% are graduates.
Currently, the group knows Box, Point, Polygon, Brush, and Auto segmentation tools and can categorize and label attributes. They are working on tasks spanning multiple use cases including retail optimization, autonomous driving and face emotion recognition.
The Pilot team is on simulation-based projects which require the application of a combination of tools, categorization, and attribute identification of brands and facial expressions of persons.
We are proud to announce that the pilot team is now going through certification and is able to meet Industry quality standards.