Hi, it’s James again and in this post I am going to explore how we intend to use artificial intelligence (AI) as a business and how you can help us to build out Pia.
Don’t believe the hype
AI is one of the most overused buzz words I have come across for a long time. At the most basic level it indicates a computer demonstrating some form of intelligence in performing a task that is normally completed by a human. Sound suitably vague for you?
The problem is that something that was considered cutting edge AI even 10 years ago would now definitely not be considered AI. For us as a team what is far more important than whether something is AI or not is whether a process can be efficiently automated to better serve our customers or partners.
Pia is our AI solution and yes, initially at least, it will act as an assistant for our customers (more on that shortly). However, it will be at the heart of everything that we do as a business. It will serve our customers, staff and even our partners from the investment industry.
The more efficient that we can be as a business the more that we can invest in to the areas of the business that need focus, such as finding and supporting new customers and developing market leading content and knowledge.
Customer first
One of the most frequently used concepts of AI is natural language processing. This is where a computer processes human language in some capacity. It is used in our daily lives, voice search with Google through to chatbots interpreting your free text responses to deal with your query. NLP technology is improving all the time and we believe that it is now at the level where, if used correctly, can form a key part of a customer journey with a brand.
Pia will be a customer assistant that will support our main site as opposed to a standalone chat interface. This will allow us to train Pia to be that much smarter given the additional interactions that the customer will have with the site. Customers can ask Pia questions about investing, whilst Pia can learn about the customers by asking them questions and understanding the actions that they are taking on the site.
The level of service that Pia can offer a customer will mature over time. We intend to move from a supportive interface to a service interface (i.e. performing a service if a customer requests it) and our long term ambition is to get to a predictive interface.
This would allow Pia to anticipate a customer’s unstated needs based on the information that it has on that customer. This target for the business will take time and ultimately we will only move to this point when the technology is mature enough but more importantly if they customers request that level of interface.
Only as smart as the people that train it
Building genuinely useful AI is incredibly complex and takes time. If you want to become a world class golfer you better make sure you are trained by someone that knows how to play golf. AI alone is no silver bullet and it needs to be nurtured and trained by experts, and in our case experts from the world of savings and investments. This is why we formed an highly experienced team and advisory board to shape Pia to become an insightful assistant for those wanting to invest.
The other challenge AI has is the need for a large number of users. This is critical as high volumes of users help to train and refine the interactions of the AI solution. Access to this volume of data is a much discussed problem for new startups in the space.
We need as many people as possible to use Pia to train it. Our Pia Community is growing every day but we would love to get more people signed up to begin using Pia when our Beta test starts in the summer. You can sign up now at yourpia.com.
There is so much more that I want to cover about our AI roadmap, particularly how it will serve other areas of the business. This is enough for now though, more to come soon.
