Progress - through change - is the only real direct measure of innovation; but at-scale change can feel daunting and the process of initiating innovation can, and most often does, create more talk than action. On one hand, the avalanche of new and emerging technologies creates hype, noise, and options that business leaders are pressured to consider. On the other hand, we live in a world of scarcity - from energy and budgets, to time and skills.
No large-scale innovation starts as such - when done right, it often begins with individuals who are willing to ‘think big’ and try something different to achieve a new and better outcome. But critically, it also requires fast action, pockets of experimentation, and an appetite to start now. So how do organisational leaders tow the line of adopting new and emergent technologies that drive innovation - balancing measured decision making to optimise and secure their solutions, while embracing a culture of failing fast?
Problem Solving through Experimentation
AI has opened up people’s imaginations and changed the way business leaders approach the same or new problems. Leveraging hackathons to experiment can be an effective way to expedite problem solving while fostering collaboration - but they must be targeted at real life problems vs. a playground of technologies. This isn’t about hackathons for hackathon’s sake - it’s about purposeful, actionable experimentation. Create a culture of experimentation and hire for problem solving skills alongside technical capability. How will you know you’ve been successful in doing this? Problem-focused experimentation is successful if their outcomes are real. In other words, what you create can be scaled up into effective production-grade solutions that meet your organisational needs.
From Experimentation to Action: Rapid Prototyping
Rapid prototyping is the next step in this actionable innovation process - taking the ideas and concepts developed during experimentation and turning them into tangible, testable solutions. A rapid prototype can prove the viability of your solution, demonstrate that the technology you’ve identified during experimentation is capable of solving your specific business problem or challenge, and create a tangible asset to test, validate, and gather real feedback against.
The reality of many emerging technologies - and in this case Generative AI - is there’s still a lot of hype, limited public references, and very little benchmark data to help organisations build the financial or business case for expansive AI investment. By focusing on producing something tangible and actionable as early in the journey as possible, demonstrate that value up front will accelerate adoption in a more meaningful way. But again, it’s not about AI for AI’s sake - by remaining focused on the business problem you will ensure the solution you build actually solves for it, rather than trying to wedge new technologies into a solution where it’s not required.
Westminster City Council: From Prototype to Production
What does this look like in reality? From the heart of London, Westminster City Council - a local authority that welcomes over 25 million tourists every year - implemented a Generative AI solution to enhance user experience and expedite internal operations for the 30,000+ waste related issues reported by residents annually. The existing solution required users to submit information manually through a web form - often creating inaccuracies, requiring redundant information, and creating a generally tedious experience for the user.
Through experimentation, Westminster City Council worked with Ancoris to iterate a solution from prototype to production in just a number of months. By starting with an initial problem statement: when residents upload a photo, it triggers a sequence of events from report submission to dispatch of the clean up crew - Ancoris and Westminster began to ideate a solution that solved for this integrated challenge. A proof of concept phase enabled the teams to prove out the potential of Generative AI and Machine Learning models - trained in Google Cloud’s Vertex AI platform for image recognition - to solve the problem. Productionisation then required a Generative AI conversational agent for user assistance, a modern and accessible frontend application to guide users through the reporting journey, integration into business systems for seamless integration, and geolocation functionality to capture report location information.
Mission Complete: Driving meaningful outcomes to solve your problem
For Westminster City Council, the goal was to create a user-centric solution that prioritised experience, simplicity, and innovation to solve a specific problem. Following the launch of the solution, report filing times are down significantly, accuracy has increased, and more residents are using the AI-enabled solution than the manual online form.
This blog post was originally published on Tech UK as part of their Building the Smarter State Week 2024.