The New Enterprise Interface

interface.ai

How do we interact with computers today? There are a lot of challenges to get things done in spite of access to immense processing power, information, and services at our disposal.

Can we optimize the way we interact with computers that would enable us to achieve more?

To further understand the problem, let’s take a deep dive and look at how humans and computers interact today.

Stages of an interaction between human and computer

There are three stages to any interaction with a computer –

  1. Intent – Determining what we want to use the computer for
  2. Discovery – Exploring to identify how the intent can be achieved
  3. Action – Providing inputs needed to ensure the intent is actioned
intent-discovery-action

Let us take a simple example to understand the stages better –

  1. Intent = Be wanting to transfer funds to a loved one.
  2. Discovery = Browsing to identify where & how to initiate a fund transfer.
  3. Action = Provide details to complete the fund transfer.

At interface.ai, we have analyzed millions of intents and we find that, on average, the time spent on each stage of interaction and based on the type of task is as follows:

We can see that users spend a significant amount of time on the ‘Discovery’ and ‘Action’ stages of an interaction, especially when it comes to getting complex tasks done.

Understanding how this transpires in enterprises and the magnitude of the impact will showcase why this is a monumental problem that needs solving.

Enterprise Impact

Enterprises spend a lot of resources to optimize Discovery & Action.

For example, sales teams are the breadwinners for any organization and it is of prime importance for an organization’s top-line that sales teams are effective – however, there are studies that show that sales reps spend almost 65% of a day doing administrative activities even with significant technology investment on tools like CRM’s. This adversely impacts the top-line of an organization.

Another example, depending on the type of industry, call centers can help generate up to 60 percent of revenues by ensuring customer satisfaction. It takes an agent around 14 weeks to be fully functional and independent. That is a lot of time to just enable ‘Discovery’ and does not seem like optimum use of time & resources. Additionally, when we also take into account the call center attrition rate of ~30%, the efforts that have gone into training is completely lost, with new agents having to be trained again.

In the US alone, the overall spending on workforce training was close to $87 billion in 2018 with spending on training payroll and products to enable & support training, growing at more than 13% year-on-year – this is an inefficient use of funds and a large impact on the bottom-line of organizations.

Looking at the automotive industry, we have seen multiple examples of recalls of vehicles due to un-intuitive interface design for ‘Discovery’ and there have been instances where this has lead to crashes and even loss of lives.

Today’s enterprise channels are inefficient leading to negative top-line, bottom-line and customer experience, sometimes even deaths. We believe with a better enterprise interface, we could create enormous enterprise value, more than that, truly enable humans to seamlessly use computers.

Market Direction

  1. Companies are attempting to solve this problem in multiple ways and there have been billions of dollars of investment made in these companies. For example, a lot of companies like Pendo and Walkme for example, are working on helping people navigate the features of a product or a service to optimize for faster ‘Discovery’.

Such efforts are incremental and do not address the core problem or create enough value for both enterprises and users.

2. We also see many being inspired by the movie ‘Her’ and building digital assistants with human-like emotions. While this is good to have, it is definitely not the first challenge to address as it does not solve the core problem. It is like spending more efforts on making a website and mobile app look pretty but not usable.

3. A few visionary organizations are attempting to build interfaces that allow for rapid action. One such organization is Elon Musk’s Neuralink, which is attempting to build a brain-computer interface that will work on enhancing the human brain (a good read on what Neuralink is aiming to achieve is available here). Through this, we could experience an intent and get it actioned instantaneously.

We predict the digital brain era will come in phases. Specifically, human to computer communication is the hardest to achieve and is close to 100x harder compared to speech recognition and language understanding in every way.

  • The first phase will enable improved quality of life for humans by combating disabilities
  • The second phase will enable better human to human communication
  • The third phase will enable better human to computer communication

Considering the complexities involved in human to computer communication through the digital brain, taking this to the larger market is several decades away.

New Enterprise Interface

Human speech has been in existence for over 5000 years now. It is our most natural choice to communicate and get things done. Thanks to Moore’s Law, and with massive affordable processing at our disposal, we can now leverage all the data that has been around to train the machines. But natural language understanding and dialogue management are still evolving and are active research areas. In the last few years, we have made good progress in making it good enough for larger market adoption. We believe the Digital Assistants which leverage voice, natural language, touch and visual all together in the right way can bring significant enterprise value and provide a big leap on digital experiences for humans. With such Digital Assistants, the interaction stages look as below

Here, the entire ‘Discovery’ phase is eliminated. The ‘Action’ phase is rapid because the interface is capable enough to take all inputs necessary at once and complete the action in no time.

Such systems are constantly learning from the data and workflows, both within and across enterprises, this leads to virtuous cycles creating highly knowledgeable Digital Assistants.

For enterprises, such Digital Assistants will lead to a significant positive impact on the top-line and bottom-line and enable customers and employees to be more efficient and get things done faster. To showcase a real-world example for this, TDECU, one of the largest credit unions in Texas is able to increase customer acquisition by 5x in just over a month since the Digital Assistant was launched. Not to forget, we are very proud to have helped many people out there get access to something very important and basic – access to financial services.

Opportunities

Currently, there is very limited data on human-computer interaction. Such interactions help us to create entirely new data sets.

These data sets will enable machines to understand human language better and enable us to welcome a new era of intelligent machines. This will enable us to infinitely scale any knowledge system and will have a multitude of applications.

For example – We will be able to infinitely scale any profession. There will be no shortage of teachers or doctors across the world as we will be able to create millions of them in an instant.

By solving for a better interface, we will be creating systems that can positively impact all aspects of human life and in turn, bring about positive changes to humanity.

The future of the enterprise is a dash and a dot.

Intelligent Virtual Assistant for Banking

Read also our blog on: interface.ai Powered Digital Assistant Increases Customer Acquisition by 5x for a Texas-Based Credit Union

AI Transformation Conversation AI Digital Assistant