Saving lives. This is the main goal of PA Consulting’s James Mucklow who is passionate about healthcare and technology. Listening to James, it is no wonder that his customer National Institute for Health Research (NIHR) is happy about their recent digital transformation. James considers digital culture as the first challenge to tackle. This, inherently, includes multi-disciplinary teams, embracing DevOps, and business appreciating the importance of IT. James’ claim that changing human behavior drives new operating models, logically follows the introduction.
Engage ESM, part of Atos, hosted a morning seminar on accelerating digital transformation, at the National Theatre. Engage ESM’s CTO Roderick De Guzman opened the event by talking about the importance of cloud and how to support it by developing and leveraging your existing ITSM organization; with a clear focus on the most-complained about factors of traditional IT: the speed and cost of IT.
The case study of NIHR was not one without its challenges. How do you collaborate with 3 million NHS staff, and implement a new service quickly? One of the solutions is what James’ team designed to help the 850,000+ people affected by dementia in the UK. The team created a site that matches people interested to take part in research and trials. The results of this solution were extraordinary, with time of recruitment dropping from months to weeks. Using the wider “going cloud” initiative, they were able to reduce operating costs by 50 %, increase productivity by 20 %, and achieve a 85 % first-time fix (FTF) rate on the self-service portal.
Other talks at the event included Chris Pope of ServiceNow, who talked enthusiastically about bots, machine learning and augmented reality (AR); possibly also inspired by the acquisition of DxContinuum by ServiceNow in December. For the more impatient, bots can give a much-desired fast response. “Bots are really just content that you can buy pre-packaged,” explained Chris. Bots can be used in e.g. 1st level support to answer queries from customers, or automatically route tickets based on their description.
According to Chris – and I would agree – the biggest problem with deploying a new AI (artificial intelligence) solution is when the problem is not understood, or proposing a solution too early. Chris gave an example of a mine that previously had a time-consuming, manual task allocation and management process, which transformed to using sensors to automatically identify valuable loads and further setting the parameters for post-processing units.
AI needs historical data – and patience. But, we need to be careful with how bots are trained; “supervised training” can easily lead to bias and discrimination1. Thus, as with any technology deployment, we should concentrate on the humans using the solution. Raising the awareness and unconscious bias amongst users of AI should be made a priority. When the true power of – any – technology is understood by its users, it will make the world a better place. And help us save lives.
- Google blog: Equality of Opportunity in Machine Learning https://research.googleblog.com/2016/10/equality-of-opportunity-in-machine.html (accessed 26th April 2017)