"Lead with purpose" say's Hyperscience's John Woodcock to leaders at Government Transformation Summit

"There is a real chance for you in this room to lead with purpose, to make your services faster, fairer and more responsive." Those were the words of John Woodcock, VP International at Hyperscience, speaking at Government Transformation Summit. 

Woodcock presented a keynote speech on "Delivering Effective Public Services Through Intelligent AI Automation." Hyperscience provide an AI driven content processing service for organisations with vast numbers of documents featuring huge amounts of data. 

For Woodcock, documents are central to government transformation. "They're the entry points to public services and the outcomes they drive are profoundly human," such as paying benefits or housing the vulnerable.

Yet government faces many challenges in extracting valuable insights from the data in these documents. These include "manual processes, outdated systems and fragmented data across government."

Additionally, Woodcock foregrounds that research estimates much of government data is "dark," meaning that it is collected but not put not analysed or used. 

While AI presents a chance to overcome the manual processes which may be holding government organisations back, if data remains dark then AI models will be less useful as they require complete and accurate data to improve processes. 

The answer, according to Hyperscience

Woodcock explains Hyperscience's answer to the challenges facing government data management, a five step approach based around a digital assistant. The organisation follows the steps below to develop a relevant AI model which will process their date by:

1. Defining the data and the processes that matter, clarifying where we need to look for data  

2. Training the digital assistant with representative examples of data usage where value can be understood  

3. Providing quality assurance by looping humans in to oversee the AI model and adapt accordingly

4. Supervising and teaching the model, increasing trust

5. Measuring and scaling the model, expanding it to more areas and use cases 

"This is what intelligent automation, at scale, looks like," says Woodcock. The digital assistant is already in place in the US Department of Veteran's Affairs where it is processing over 1 billion pages of documents each year, saving the organisation over $470 million. 

As Woodcock highlights the key is "using Ai to unlock the value of the data that you actually already have" in a government environment characterised by rising demand, tight budgets and the requirement to modernise.