The Journey of V-Soft Digital AI Evolution

Konrad Konarski, V-Soft's AI & IoT Practice head, shared the evolution of V-Soft's AI technologies and offerings with end-users and clients at the recent Big Data & AI Toronto virtual event. This process has allowed V-Soft to support businesses in evolving their own AI journeys. Check out the video above for Konrad's full presentation or catch the recap below. 

V-Soft's Chatbot Enlightenment 

According to Konrad, "User perception and authentic interaction are two key parts to a successful adoption of a chatbot, regardless of how precise it is in responding to information or how well it is implemented." This point is illustrated in the below image. While "chatbot" is in the centre of the image, perception of chatbots and the use of analogous technologies sits outside of that target. For example, Computer Vision and NLP, the fundamental building blocks of a chatbot, sit within a 1-3 year limitation.

What does this mean? A lot of people are implementing chatbots, but their understanding of what Natural Language Processing (NLP) is, is not there yet. Why is there a disconnect? Why is AI evolution happening and what is it accomplishing for our communities of customers?

In his presentation, Konrad touched on 2 examples of the AI evolution. 

Evolution of the VERA Chatbot

As a way to explain how a chatbot can be enhanced over time, Konrad shared a 3-year journey of VERA, V-Soft's AI-powered chatbot.

• 2018 - Simple chatbot built on cloud services stack
  The issue with this black box solution? Customer's don't always want a cloud-based solution. Additionally, the ability to custom train models and make them more efficient is next to impossible. 
• 2019 - Sophisticated virtual agent, decoupled NLU and domain-specific 
  The technology platform was built on an open-source NLU model and was pre-trained. This allowed the NLU engine to sit on-premise or off-premise, which gave the flexibility to build custom, visual dashboards and made VERA an independent chatbot platform - beneficial for end-users. 
• 2020 - Focused on user experience, harnessing knowledge This year V-Soft focused on the core adoption driver for chatbots - the ability to interact with a virtual agent in the same way or better than a human agent. Year 3 has to lead to a more interactive and compelling chatbot, maximizing the adoption rate of this technology. 

The Evolution of Computer Vision Technology

Konrad also went into the 3-year journey of Computer Vision within V-Soft's offerings, as this technology is becoming a part of many organizations. 

• 2018 - Custom models and data insights 
  At its basic level, Computer Vision technology consumes video footage and provides charts and analytics into the information from the footage.
• 2019 - From platform to 3rd party integration
  As this technology evolved, there was a realization that there many vision platforms on the market. V-Soft instead built a platform that can ingrate into existing application systems. 
• 2020 - Edge AI and value-driven 
  Much of 2020 has focused on Edge AI, with the ability to analyze Computer Vision feeds and provide actionable decision making on the spot. Some people don’t have the bandwidth to run Computer Vision feeds into a cloud, so edge-based camera systems are invaluable.

Closing the Gap on the Disconnect

While technologies within AI have progressed and evolved over time, it's important to close the gap on the disconnect of AI's perception. This disconnect occurs from top to bottom, from upper management down to IT. Stakeholders may disagree on the feasibility of implementing AI technology. Others may not consider the broader strategic efforts of AI. If we bridge this gap, the power that AI can bring to the business can be truly remarkable and will continue to evolve over time.