Intelligence-Led Regulation with Objective RegWorks

Realising new opportunities in intelligence-led regulation through a multi-disciplinary approach

This webinar features John Munro, a Regulatory Practice Specialist with extensive experience designing and leading national data, insights and intelligence functions across government organisations in New Zealand and the UK.

In this session, John shares a compelling real-life case study on leveraging organisational data to implement intelligence-led regulation, in turn unlocking new opportunities through a multi-disciplinary approach. 

You can also read John's thoughts in our latest insight paper here.

Additionally, the Objective team showcase some key features of Objective RegWorks, demonstrating how it facilitates data-informed regulation, enhances decision-making and harnesses the data held within the organisation.

Highlights include:

  • Visual Associations: A comprehensive 360-degree view of entities, including their historical and current connections with other organisations, individuals, and locations, alongside detailed contact and case histories.

  • Risk-Based Analysis: Utilising risk-based algorithms that can be tailored to specific parameters and weightings, enabling the calculation of precise risk ratings.

  • Offence Management: Easily configurable for changing legislation and penalties, facilitating precise enforcement. Data can be utilised for trendspotting, informing policy changes, and enhancing proactive regulatory approaches.

  • Dashboards: Providing rich visualisation tools for effective workload management and KPI monitoring. Prioritise tasks, delve into individual records, and generate real-time operational reports from customisable personal dashboards, tailored to working preferences.


Audience Q&A
This section provides John's written responses to all webinar questions.

Q1. What's the difference between intelligence and insights?

In principle, intelligence has a narrow focus compared with insights – delivering a regulator’s strategy is driven by insights and other key factors.

There are some questions that can’t be answered by evidence-based methodologies such as research, evaluation, business intelligence and advanced analytics. This might be due to insufficient resourcing or poor information quality. Intelligence analysis aims to fill this gap – supporting decision-making under uncertainty. Intelligence draws on every available source of information to develop insights. This can include data, but also includes debriefs of staff, industry experts, researchers, workers, and victims, research papers, industry press, transactional records, and information held by other agencies. All these sources are evaluated for reliability and drawn together like pieces of a jigsaw puzzle. Probabilistic language is used to communicate the degree of uncertainty associated with the insights.

Insights are produced by consistently following a process in which information of all forms – from data to personal experience – are collected, processed, and analysed to produce actionable insights: and then used to impact our environment. There are a range of methodologies available that can be used to process information, such as intelligence analysis, advanced analytics, research analysis, and evaluation. The type of methodology best applied to an insights-gap depends on several factors such as the amount of funding available to invest in insights-creation, the degree of certainty required for the decision, and the quality of information available for analysis.

Q2. There has been a lot of talk about 'democratising AI/Machine Learning' so that end users can do this sort of work on the fly. Do you see this as being a common direction and are you aware of this type of automation happening?

I see parts of this being realistic, and parts are still kind of fantasy thinking. In many cases, this is the same as saying ‘democratise statistical analysis’ which immediately makes it clear how plausible the endeavour is. It can work where tools tailored to specific decision types are created by data scientists and then put into a package that’s ‘foolproof’. As an example, there was a model that was built by my previous team where the operational arm was provided with a lot more information, that supported their decision making, but did not need to be built by themselves.

On the AI front – to some extent, it’s now a different ball game. There are an increasing number of easy-to-use, large language models that can produce remarkable results – with caveats. Even the best of them will produce incorrect results – and you have to know enough to be able to detect these. So, expertise is still valuable. However, they will increasingly represent the ‘front end’ through which people interact with suites of other AI/data science tools. Over the next few years, I suspect this will become the way that most people leverage data and models. The LLM will frame the request and assign it out to sub-models – eventually producing a result. Early experiments putting tables of data into Chat GPT and asking for basic analysis, produced some pretty remarkable results. Ensemble models will make these early results look ordinary.

Q3. What would you see as the necessary disciplines within an insights team within a regulatory agency?

  • Futures thinking – Having a degree of passivity in that insights teams will produce products on request but aren’t necessarily scanning for risks and opportunities and bringing them back to the organisation, is a risk for the team and organisation. Waiting for management teams and ELT to direct their work is not the best approach, be proactive rather than reactive ... apply a forward-looking lens.
  • Understand the internal decision-making landscape – Consider what factors go into a good decision, as well as what the most important decisions the organisation needs support with are and the timing of them.
  • Visible leadership that influences change is important. Products are produced that the organisation does not pick up and use. That can be down to a number of factors, including, but not limited to, a lack of organisational understanding of what insight teams actually do, poor commissioning and prioritisation processes. Being positioned to educate and influence Exec Leadership Teams is important to reduce this type of ineffective/inefficient churn.
  • Have the ability to weight and evaluate different sources of information – as part of the information synthesis process.
  • Some others to consider – data literacy, succinct writing skills, data visualisation, data modelling, statistical literacy, AI/Advanced analytics (if they sit within a team) and critical thinking.


Q4. There is a book by Russell and Hodges called 'Regulatory Delivery' and he suggests a ratio of 3:1 for operational frontline staff to analytics staff. Sparrow also talks about having a 'Nigel'. What are your thoughts on that ratio and how to best manage the ability to deliver for the organisation?

Having a “Nigel” is a valuable asset, as they draw sense and interpret the amount of data/information floating around. However, to be effective and efficient, a Nigel needs a manager/leader who “gets it” and can drive it in a way that creates an appetite for it from the CE down. The Nigel/Manager relationship is important, or the “Nigel” becomes frustrated and decides to move on.

As AI develops and insight productivity grows, there is the potential to increase the ratio. To make this viable, it is important to develop a deeper understanding of the parts of the analytical process so that the non-value add parts can be automated. This requires the more technical people having a link into the insight creator process – and insight creators need to have a bit of an understanding of what roles such as data scientists are up to – so that that iterative automation process can happen.

The goal of that process is not about replacing analysts, more about being focused on taking the drudgery out of their workday and letting them focus on getting to the ‘aha!’ moments that they’re in the job for.


Presenters:

John Munro

Regulatory Practice Specialist, Objective

john.munro@objective.com 

Mat Graves

Regional Director (UK), Objective

mat.graves@objective.com

Jason Riches

Solutions Consultant, Objective

jason.riches@objective.com