Online accounting provider Xero’s recently announced plans to incorporate artificial intelligence into its systems is further proof that far from being a future technology, AI is already a business-ready tool.
As mentioned in my previous blog about the 2013 CIO Summit, I believe AI can help make online and cloud-based tools more responsive to user needs. I even suggested it could be the “fourth platform” after the so-called “third platform” created by the convergence of cloud, mobility, big data and social media.
By adding AI to its tools Xero aims to help streamline labour-intensive processes for its clients – particularly accounting firms.
It is using AI technology provider Celaton’s inSTREAM product, which automatically extracts key data from invoices, receipts and financial documents received by email, fax, post and paper into the Xero system.
This is a great example of how AI can enhance online and cloud-based tools by taking care of routine tasks with less intervention from users who are then freed up to focus on activities that deliver value to a business and its clients.
There are already many mainstream examples of how AI is making cloud tools more responsive to user needs.
Many smartphone users will be very familiar with voice recognition tools like Apple’s Siri and Google’s Voice, which learn and adapt from their users. We also see AI at work in online translation services like Google Translate and location-based services that offer specific content or information based on where you are.
The power of these tools is that they use natural language to process questions or requests and to provide answers – so you don’t have to learn to talk in code language!
Neural networks, which power optical character recognition used for business card reading apps, are another example of AI.
In future we could see cloud-based expert systems which people can use for self-diagnosis of health issues. Being able to describe your health issue to a machine like you would to a human doctor is where AI comes into its own.
But what truly sets AI-enabled tools apart is their ability to learn and improve. Through the concept of ‘machine learning’ systems learn from users’ preferences or behaviour to improve how they work.
We see this in services like Amazon or iTunes where purchase suggestions are offered based on user history or with Spotify where music suggestions are made based on what you’ve listened to before.
This learning is enabled by genetic algorithms which mimic the process of natural evolution, as well as by optimisation code that modifies systems to make them work more efficiently.
At Psoda we currently use AI for a range of tasks in our online management tools, such as calculating critical paths in scheduling, keeping track of user preferences and highlighting critical information for example overdue tasks and milestones.
In the future we want to expand the application of AI in the Psoda suite of portfolio, programme and project management tools. For example creating report templates from natural language descriptions, location based services, machine learning and data analytics.