GOVIS Events - 2025

Lunchtime forum - Artificial Intelligence in Testing

Mon, 19th May 2025, 12-1pm, G.03, Ministry of Business, Innovation & Employment, 15 Stout Street Wellington 6011 - and online

AI in Testing and Testing AI – The Hype is Real 😊

AI is looking to be crossing the threshold from being talked about to being used within the testing domain in Government.

As part of the regular engagement between GOVIS and the Government Information Technology community, our most recent GOVIS Software Testing Professionals Meetup was held online and in person. We were honoured to have presentations from the GCDO team with the Department of Internal Affairs and TTC Global New Zealand, as well an announce the results of a brief (and unscientific) survey on the use of AI across those that registered for the event. With over 150 people registered for the event, this clearly was a topic of real interest to the GOVIS Software Test Professional community.

Our first speaker – Pedro Ramirez – set the scene with a really useful summary of the state of the Public Service AI work programme and Usage across Government. From the presentation it was quite clear that the New Zealand Government has a measured and well thought out approach to the adoption and use of AI, focusing on the themes of Enabling (making it easier for Government to use AI), Safe (establishing fundamental guardrails for AI use) and Responsible (ensuring that the public’s trust is maintained in using AI). Pedro talked through the different initiatives across the three themes examples being AI Community of Practice, AI Toolbox (including a Māori perspective), Responsible AI Guidance for the Public Service and AI engagement at a domestic and international level. The presentation was highly interactive, well-paced and well appreciated by the attendees.

Our second speaker – Mei Reyes-Tsai from TTC Global New Zealand then talked to the practical approaches and issues faced for the use of AI in Testing and Testing AI. Mei’s presentation talked to the challenges and risks of AI adoption – as well as addressing the hype associated with AI in the software testing domain. Use Cases for the use of AI in Software Testing were covered and rated. This is a programme of work that TTC Global is updating at the AI environment continues to evolve and mature. Mei also presented on the risk areas of AI implementations by organisations – highlighting some real-world incidents. And then also walked the attendees through the difficulties in testing AI-based implementations (non-deterministic responses, the emergent behaviour of AI), how that impacts current testing strategies and presented on innovations on testing practices to mitigate these issues. Overall, a useful assessment of both using AI in Testing and Testing AI – one that can be picked up and used as a blueprint for individual organisational implementation of AI.

At this stage – given the active Q&A – we were over time for the event. A sign of a well-engaged audience!! The last session was a quick presentation on the survey results. Key takeaways were:

·       At an individual level, people were using AI in their work activity.

·       Organisationally, agencies were in the planning stage of AI use, with limited / early-stage AI implementations.

·       Overall people saw the productivity increase potential of AI in the workplace – yet were still wary of the impacts of AI on the workforce.

Most concerning was the question the movie Terminator being the source of their AI information. While (thankfully) a very limited number said that the Terminator movie was their AI reference point, the survey did report that a number of respondents had not seen the Terminator movie. Easily the most shocking and disturbing find of the whole meetup!!

For reference and further information, the event presentations are attached below.

The GCDO AI Work Programme presentation is available on email request for the meetup attendees. It has not yet been approved for general publication.

Lunchtime forum - A case study in digital government from the USA and Canada

Wednesday, 7th May 2025

We heard from Aaron Snow – one of the world’s 20 most influential people in digital government.

This presentation was part of a series called Transforming Government - where GOVIS and our partner Silverstripe take a look at some of the ways that digital technology can enable positive change in the way that government operates and delivers services.

We heard from Aaron Snow, who specialises in helping governments serve people better. Aaron has led digital service teams for the national governments of the United States (where he was Deputy Commissioner of the Technology Transformation Service) and Canada (where he was the first CEO of the Canadian Digital Service). Aaron has been named one of the world’s 20 most influential people in digital government. Aaron is currently an advisor with AWS, helping governments around the world on their digital transformation journeys.

Aaron joined us live, online from the United States to share his reflections, insights and predictions for digital government. He spoke in a personal capacity, and his comments do not represent the views of his employers past and present. Here are a some of his general observations:

  • It's really important to not just blindly accept received wisdom and rules, but to challenge these and try to get them changed - if it is getting in the way of delivering value to people.

  • Maslow's hierarchy of needs can be applied to the public service - the most fundamental thing is legal security (i.e. not acting illegally), then political security (i.e. not getting in trouble with politicians), then bureaucratic security (i.e. not annoying your colleagues), then being able to perform the bare basics of your job, and then finally at the top of the hierarchy you have improving the way you do your job (continuous improvement). So if you want innovation from the public service, then these other things need to be in place first. 

  • Conway's law tends to be true - this can be paraphrased as "show me your org chart, and I'll show you your strategy!" One implication of this is having a central government digital design authority is key for making progress in digital transformation of government - otherwise many projects will end up failing, and the user experience will be confusing and fragmented.

  • Apps are increasingly important as more and more people use mobile phones for accessing digital services. One important goal for a central government digital design authority should be to offer a single mobile app for accessing government services.

  • Make change a performance goal for senior leaders - for example how much automated testing do they do? How rapidly do they deploy new products or features? How many of their services have API coverage?

  • Outcome risk is more important than project risk. Disasters like Phoenix payroll in Canada and the Columbia space shuttle can be attributed to a narrow focus on project risk.

  • Remember that the public service technology we are building is part of a long game, where we are enabling venerable government institutions to transition to the digital age and continue underpinning our societies for the next hundred years or more.

And here are some of his practical tips for anyone trying to innovate:

  • Docs or it didn't happen - be sure to document everything, so you can refine, make decisions, and move on. Documentation means you can't be ignored, and that a discarded idea can resurface when its moment arrives.

  • Remember that code is cheap, be prepared to tear things down as readily as you build them (so you don't get stuck with something mediocre).

  • Use demos to encourage the art of the possible (a demo is worth a million words).

  • Find good examples to copy, most public service innovation challenges have already been solved somewhere.

  • Show the whole iceberg of detail of what you did - the good bits and the ugly bits (or less no one else can learn from it).

  • A blameless retro culture is key - make a regular habit of talking about what has worked and what hasn't in a way that is about outcomes and not individual performance. Doing it often will keep the stakes low and make sharing easier.

  • Remember that being innovative doesn't necessarily require too much bravery. Really, it is more about having clarity and always being honest. If you just focus on articulating the truth, then you don't need to worry too much about the rest of it.

Lunchtime forum - Unlock the power of data with the Unified Star Schema

This event took place on Mon, 24th March 2025. You can watch the presentation here on YouTube. Our thanks to Astrato and OSS Group for sharing their event with GOVIS

The USS is an innovative platform-agnostic approach to data modelling, that makes it easier to access all your data products in one place.

Are you a data architect/engineer/scientist/warehouse developer? Are you that person who always finds themselves saying "what about the schema?" Do you wish it was easier for your BI customers to 'self serve'? Do you wonder about how to leverage genAI tools for data analysis? Then you may be interested in this presentation...

The evolution of AI is significantly reshaping the way organisations are looking to utilise data, which brings new data modelling challenges. While many tools offer appealing data visualisation techniques, users can still often face difficulties when seeking broader insights.

The Unified Star Schema (USS) aims to tackle this challenge using an innovative data modelling approach. It's designed to enable seamless access to all data products through a single access point, regardless of technology or platform.

In presentation we learned about:

  • Options to prepare for the impact of AI assistants: As AI develops, it is vital to enable direct access to data for business stakeholders.

  • How stakeholders benefit from simplifying data analysis and reporting across various products.

  • How to reduce the workload associated with building and maintaining complex reports and dashboards.

Speakers:

  • Francesco Puppini, inventor of the Unified Star Schema and the Puppini Bridge. Published a book about on this topic in 2020 here, that was co-authored with legendary 'father of the data warehouse' Bill Inmon.

  • Konrad Matheis, COO / CTO of Astrato.

Lunchtime forum - Transforming Government through System Leadership

Fri, 14th February 2025, 12-1pm, G.03, Ministry of Business, Innovation & Employment, 15 Stout Street Wellington 6011 - and online

From GOVIS, in partnership with Silverstripe

It was great to hear from Monica Greenan (DIA - General Manager, Agency Partnerships & Capability, Digital Public Service) about some of the many different areas where the GCDO are active, for example:

  • Taking steps to further centralise digital procurement.

  • Collecting insights into NZers' (extensive) use of digital channels.

  • Rationalising the number of digital identities NZers need to interact with government (e.g. ACC's use of 'My Health Account Workforce') and continuing to progress a 'growing ecosystem of digital identity' under the Digital Identity Services Trust Framework Act 2023 (DISTF).

  • Weighing up the pros and cons of 'sweating' our on-prem IT assets.

  • Emphasising the impossibility of all agencies getting funding for their digital investment wishlists, and the need for more shared/centralised capabilities (e.g. customer portals, CRMs, case management systems, records management systems, registers...). This is the driver behind the Service Modernisation Roadmap - which highlights the joined-up work that is happening (or ought to happen).

  • Reminding agencies that we are all responsible for pursing shared solutions - e.g. if you at the point of spending six weeks talking to vendors as part of procurement market investigations, you should also be speaking to other agencies and the GCDO to see whether there might be a way to save money through collaboration.

  • Advocating with Treasury to make it easier for agencies fund IT investments through opex, not just capex.

  • Exploring the opportunities presented by Digital Public Infrastructure (imagine if we could achieve theoretical savings of 1% of GDP!)

  • Lots of AI stuff, including understanding most common agency use cases (e.g. contact center optimisation, CRM enhancements, digital process automation), the high proportion of 'internal use' tools, and the relatively small number that ended up taken through to BAU (15 from 138), for a range of reasons. Also gained some more nuanced understanding of risks and advocating for agencies to proactively publish details of their use of AI (trust through transparency). Finally, there is also an AI Community of Practice, that public servants may request to join.


Here are some key links and resources: