15 May 2024
by Varun Sarin

What to consider when scaling AI in the public sector – Part 2

Guest blog from Varun Sarin at Credera as part of our #UnleashInnovation campaign week.

In March 2024, the National Audit Office (NAO) reported that 70% of surveyed government bodies are either piloting or planning the use of AI, with applications ranging from supporting operational decision-making to enhancing internal processes. 

In the second instalment of our exploration into AI implementation in the public sector, we delve deeper into key factors essential for scaling AI. Here we’ll cover the last three foundational capabilities: Delivery Model, Talent, and Adoption & Scaling and their associated guardrails. 

You can find the first part of the blog here: What to consider when scaling AI in the public sector (techuk.org) 

Let’s start by looking at the Delivery Model.  

Unlike traditional projects, AI solutions often necessitate a flexible and experimental approach, making them akin to Research and Development (R&D), demanding flexibility and experimentation. Old-school delivery models don’t work because they don’t require a re-design as often as AI solution delivery currently does. 

The essence of AI lies in its adaptability and the ability to pivot swiftly in response to evolving requirements. Locking yourself into one rigid delivery model can create significant challenges down the road.  

Agile methodologies, characterised by their iterative and time-boxed approach, emerge as the preferred approach in this scenario. According to Forbes, these methodologies facilitate rapid experimentation: it's all about testing, gathering feedback, and delivering value continuously.  

A fundamental consideration in the delivery model is the approach towards tooling and automation. Establishing robust Continuous Integration/Continuous Delivery (CI/CD) pipelines is important to ensure seamless deployment and management of AI artifacts throughout their lifecycle. Frameworks such as the Llama Index offer a pragmatic solution to invoke AI tools, obviating the need to reinvent the wheel. By leveraging existing frameworks, organisations can accelerate development cycles and mitigate the risk of vendor lock-in.  

However, with innovation comes inherent risks, so how do you protect against these risks? Strategies such as progressive release of capabilities and meticulous monitoring; A/B testing, model monitoring, and human-in-the-loop mechanisms emerge as indispensable safeguards against unforeseen anomalies and biases.  

During the build phase, leveraging tools like tools like DeepEval can be a valuable  lifeline They help assess model relevancy and spot aberrations like hallucinations and biases.  

Now, let’s move onto Talent. 

The National Audit Office highlighted that 7 out of 10 government bodies cite skills as a significant barrier to AI adoption. Addressing this challenge requires comprehensive training across all roles. But here's the harsh truth: there's a dire scarcity of AI talent. With roles like those at Netflix commanding salaries up to $900,000 annually, public sector budgets simply can't keep pace. So, rather than chasing unicorns, organisations are doubling down on upskilling existing teams to optimise off-the-shelf AI solutions. It's a pragmatic pivot given the constrained resources at hand. 

At Credera, we have developed tailored training pathways catering to both technical and non-technical staff. For technical roles such as cloud engineers, dedicated programs focus on AWS, Azure, or GCP, with a stringent training target set at 95% for the organisation. Non-technical staff undergo a combination of broad skill development, familiarisation with AI terminology, and interactive learning challenges to level up their game. 

Aligning the AI journey with the interests and responsibilities of different departments is crucial. Think Legal, HR and Marketing. Plus, don't forget to prioritise training for new hires and graduates, as they have the most risk of losing early experience because of the introduction of AI. 

Measuring the AI team's proficiency goes beyond technical skills. It's about ensuring they understand compliance, risk mitigation, safeguarding sensitive data, navigating intellectual property considerations, and mitigating biases in models.  

Finally, let’s address Adoption & Scaling. 

According to Harvard Business Review, over 70% of large companies are still grappling with integrating AI effectively. As organisations scale their AI initiatives, the pressure to deliver value regularly intensifies. With expansion comes the need for consistent and tangible outcomes to justify the investment in AI technology.   

Getting everyone on board is key. Think broad adoption across departments. Transparency is crucial too, helping stakeholders differentiate successful initiatives from those needing a reboot.  

At Credera, we love celebrating wins. From training uptake to revenue bumps, our success metrics cover it all. But it's not just about shouting from the rooftops. Success stories need to reach every part of the organisation. And who better to lead the charge than the executives themselves?   

Central to fostering widespread participation is embracing the experimental nature of AI integration and cultivating a culture where failure is viewed as a stepping stone to innovation.  Moreover, recognising and measuring usage across the organisation not only justifies investment but also fosters accountability and ownership. By making AI utilisation visible and quantifiable, organisations can garner support and enthusiasm for further integration initiatives.  

Just as in any technological evolution, success hinges not only on the sophistication of the tools but on the commitment and adaptability of the individuals and organisations wielding them. With a steadfast commitment to innovation and collaboration, the public sector can harness the transformative power of AI to drive efficiency, equity, and excellence in service delivery, thereby ushering in a new era of public sector innovation and advancement.


 View all campaign week content 

Unleashing Innovation campaign week graphic.png

techUK – Unleashing UK Tech and Innovation 

innovation_icon_badge_final.png

The UK is home to emerging technologies that have the power to revolutionise entire industries. From quantum to semiconductors; from gaming to the New Space Economy, they all have the unique opportunity to help prepare for what comes next.

techUK members lead the development of these technologies. Together we are working with Government and other stakeholders to address tech innovation priorities and build an innovation ecosystem that will benefit people, society, economy and the planet - and unleash the UK as a global leader in tech and innovation.

For more information, or to get in touch, please visit our Innovation Hub and click ‘contact us’. 


Tech and Innovation Summit, 19 June

Emerging technologies will be debated and explored at our annual Tech and Innovation Summit, taking place on 19 June. This campaign week will directly feed into the themes discussed at the Summit. 

techUK’s flagship Tech and Innovation Summit returns to traverse the extraordinary and ground-breaking discoveries made possible by the application of emerging and transformative technologies. 

Secure your ticket now.


Upcoming events:


Latest news and insights:


Get our tech and innovation insights straight to your inbox

Sign-up to get the latest updates and opportunities from our Technology and Innovation and AI programmes.


Learn more about our Unleashing Innovation campaign:

Unleashing the Potential of UK Tech and Innovation.jpg

 


Laura Foster

Laura Foster

Head of Technology and Innovation, techUK

Rory Daniels

Rory Daniels

Programme Manager, Emerging Technologies

Tess Buckley

Tess Buckley

Programme Manager - Digital Ethics and AI Safety, techUK

Usman Ikhlaq

Usman Ikhlaq

Programme Manager - Artificial Intelligence, techUK

 

Related topics

Authors

Varun Sarin

Managing Consultant, Credera