Getting started with AI, embracing iterative progress & conquering the Frankenstack"
"How can I get started with AI? It's a bit daunting and confusing."
This question was asked in a recent webinar – Transforming citizen experiences. Where we dug into how public sector organizations can become more agile. Focussing on:
- Digital trends shaping citizen expectations
- How to construct an agile tech stack to empower rapid iteration
- Examples of organizations already adopting this approach
Getting back to the question, Anthony Nigro – Squiz GM, stressed the importance of remembering that...
"AI is a tool to be able to deliver digital experiences. AI itself is not the objective".
Anthony highlighted the importance of starting with a strong use case that can leverage AI.
So the idea of just putting on a chatbot, yes, that will assist... but really, what is that chatbot or agent there to help with?
Another thing to consider is making sure you select the right tech. AI could introduce some complexities for your organization. For example, large language models sit underneath the AI tech you're using. Depending on which you choose, there could be a cost and quality of results implication.
Above all, start with a use case and then focus on using the right technology approach that will support that use case.
Recommended by LinkedIn
Clarity of outcome vs certainty of output
In this DX guide, we propose that the future is dead. A pretty alarming statement – but necessary to highlight the extreme pace of change and the challenge of long-term planning.
Tech is evolving at speed, major market shifts occur frequently and customer expectations are accelerating rapidly. The digital experiences we offer have to keep up.
Traditional long-term strategies driven by lengthy deployment roadmaps are no longer an option.
A new approach
In an AMA session, John-Paul Syriatowicz – Squiz Co-Founder, spoke on the importance of embracing a more iterative approach with ‘success plans’.
To connect your North Star – what you're trying to achieve – through to action on a relatively short time horizon in a way that provides flexibility for change.
Develop 6, 12, and 24-month plans – but approach each phase differently.
Your 6-month plan will have the most clarity and certainty. Your 12-month plan will see you loosely fix goals. Finally, your 24-month plan will be informed by the data gathered from your 6 and 12-month activity. This allows for flexibility and changes to your output.
Chris Howard – Managing Director of Data Analytics and AI at Slalom , describes this approach as opting for ‘clarity of outcome over certainty of output’.
Tame the Frankenstack
The rise in tech has probably left you with a bloated, messy martech stack. The dreaded Frankenstack.
A monster comprised a tech that doesn't work well together, loosely tied with spaghetti integrations. These Frankenstacks drain your money, time, and sanity. They also create data siloes, security issues and make it really hard to justify ROI.
It's time to assess your martech stack – looking at what is bringing you value and removing those that are not.
"But, we have so many tools. Where would we even start?" – a common question.
Your mind probably jumps straight to the elephant in the room, re-platforming. But, fear not.
In this AMA session, Samantha Goble from XCentium , discusses assessing your tech iteratively. Allowing you to reduce costs and prioritize essential solutions that deliver consistent data and ROI.