In an era where the use of new tech is taking increasing by leaps and bounds, education isn’t far behind. One such booming startup is Skilldom, which brings the best of the new age learning solutions, creating a perfect amalgamation of the human touch, and technology interface. It is an end-to-end learning and development consulting firm which empowers the global learning community using a fine blend of outcome-driven learning strategies, technology and innovation.
Skilldom recently launched India’s first Assisted Learning Intelligence (ALI) platform iCOG in a pilot project at India’s largest manufacturing, technology, engineering and financial conglomerate in India. The platform allows for talent empowerment and self-learning by personalisation of the experience, was launched earlier in August 2017 in Mumbai and other regions. iCOG is an intelligent system that assists organisations build talent, capability and productivity in a quick and effective way.
Analytics India Magazine interacted with Anuroop Iyengar, the brain behind iCOG, who currently serves as the CTO and drives innovation and product roadmaps for three research groups in speech recognition, computer vision and text analytics application. With over 15 years of experience, he worked as an analog full chip engineer in his previous avatar, which equipped him with skills in deep math and system engineering skills.
“iCOG allows employees not only to take charge of their personal development but also aligns them to the business goals of the company empowering them to fast track their career. The intelligent system brings to employees the learning that helps them transcend to conventional learning boards”, says Iyengar.
Currently, India’s largest manufacturing, technology, engineering and financial conglomerate in India is embracing iCOG. It is currently being rolled out to 35,000 employees out of the 95,000 employees wherein the employee engagement and capability levels are tracked on a real-time basis.
Technology Behind The Intelligent System
iCOG extensively makes use of machine learning and recommendation engine. Its recommendation engine is an ensemble (hybrid) model that caters to both organisational goals from the learner (competency development) as well as learner’s personal interests and preferences.
“In the era of deluge of information available to users through various channels in various formats, iCOG attempts to take a guided approach for each learner with the dose of content recommendation to master his/her competencies. iCOG aggregates digital content from web, partners and enterprises, and uses statistical ML and Video AI to create a granular intelligent content index”, said Iyengar.
He further added that their proprietary patent pending algorithm creates a learning Index — HGV (‘has-grasp of’ value). The system leverages this learning index and content difficulty level to match learner’s current competencies.
“Once the profile is mapped to the target content repository, engine employs further personalisation. This is achieved by a collaborative approach to extrapolate user’s personal interests/preferences, learning styles, correlated with the activities by the rest of the Peer learners including content popularity and affinity patterns”, Iyengar said. iCOG also does a Social Network Analysis around learning activities providing an extended user experience. Their recommendation engine predicts most apt set of content and SMEs that will most likely relate to learner’s context and influence a learner.
Machine vision techniques are also being used to intelligently create bit sized learning snippets from existing videos.
How AI, NLP And ML Are Used To Recommend A Spectrum Of Learning Options
Iyengar shares that as enterprises and their businesses evolve, learners have to be continuously upskilled. “A solution that can solve this problem should be in a position to understand a learner’s context – learning styles, learning patterns, current needs from a work perspective and many other aspects that define the context in which a learner needs information”, he said. Machine learning techniques are thus used to build an adaptive learning ecosystem that identifies the best of the information made available to the learner to build an adaptive learning ecosystem that identifies the best of the information made available to the learner.
They are currently building a chatbot and a voice-enabled virtual coach application that uses deep speech techniques. This virtual coach is a dialog system for users to converse with iCOG for their learning needs.
Transforming The Way Learning Industry Works
Iyengar is quick to add that if you look at any enterprise in its current state, there are three generations of learners with different learning paradigms which means that source of learning cannot be of a single dimension. He believes that going a little into the future, most organisations will have millennial and it would be important that their learning paradigms be addressed.
Below are some statistics:
- Average learning time for an individual is around 24 minutes a week
- Time span for catching the attention of people is between 5-10
- Video content less than four minutes’ duration is what keeps learners
- There is no holistic measure of learning at an individual and
All of this means that solutions will have to become learner centric, should support open learning frameworks and encourage just-in-time learning. “Our competency measurement system, considers 21 dimensions to measure learning index,” he adds. Iyengar says that iCOG uses the learning index to:
- Enable learners to identify career paths
- To better execute succession planning and other critical functions
- Map success of L&D functions in terms of business impact
The Eureka Moment
“It came to my mind when I was engaged in a stint teaching kids in Government school in Bengaluru”, said Iyengar. The need was for a sustainable mentoring and teaching mechanism right in the middle of an ecosystem that has a very poor student-teacher ratio. “The case was no different in a corporate learning scenario where ‘employees to trainer’ ratio is not essentially healthy; it is not sustainable to increase the ratio by increasing number of trainers either” said Iyengar.
“Therefore, what we identified as a solution, is a system that created a ‘personalized tutor’ experience”. And hence came the thought of creating a machine learning solution that can automated the learning process. “What I did at the school was replicated with the machine being trained to react to every trigger of a learner”, he said. It is worth noting that iCOG is supported by a strong team comprising of data scientists, product engineers, content development and curation team that is behind building this intelligent system.
Growth Story And Challenges
iCOG, which is can be currently availed in “Enterprise” and “SaaS” models, is providing a platform for users to create and share content. As Iyengar says ‘If you want to learn something teach it”. iCOG creates a framework for every learner to contribute to Enterprise knowledge.
It cannot be denied that iCOG is a path-breaking new approach to the way online training will be delivered. “Organisations are looking at creating a personalised experience and continuous learning culture for their millennial employees and our Assisted Learning Intelligence (A.L.I) platform measures the learning and empowers the organization with data driven actionable insights for the skill development of their employees”, he said. It therefore creates a roadmap for future growth.
“We have customized ALI platform for India’s largest manufacturing and engineering conglomerate which is scalable and cost-effective. We have enquiries from across corporates in India, Middle East and SEAsia and are glad they want to be a part of the change and accepting our innovation”, he added.
However, they are not averse to challenges. Some of the major challenges in the adoption of their ideas is difficulty in phasing out existing learning systems, cultural mind-set in enterprises and existing IT policies limiting the utility of open web content.
“The HR businesses have added considerable measures in terms of execution, learning and development last year. This year, its about witnessing another huge inner move, however this time the attention would be towards developing innovative learning platforms—how is it utilized to discover individuals, interface individuals, draw in individuals, even supplant individuals. Will we be able to create a blended learning experience for the employees? The coming year will witness disruptive technology being implemented by various conglomerates globally”, he said in the concluding.
Try deep learning using MATLAB