How to successfully implement an AI system | Computing

Trendy news on Technology


While we’re seeing rapid investment in intelligent interfaces, organisational and technical challenges during deployment can see many companies struggling to transform themselves into AI-driven enterprises. Martin Linstrom, Managing Director, UK&I at IPsoft, whose human-like digital AI colleague, Amelia has the ability to learn and improve over time, explores his 5 best practices for successful deployments.

Where do you start when introducing an AI into a business?

The most important thing to remember when beginning an AI deployment is what the business outcome should be. This should remain at the heart of the decision making throughout the journey. The best and most effective way to achieve this is to ask your workforce how AI can make their everyday work easier and more interesting, ultimately resulting in the best possible user experience for the employees.

How much money should companies invest initially?

Companies should calculate the anticipated cost savings that would be gained with a successful AI deployment, using that as a starting point for investment so that costs of errors or short falls on expectations are minimised if they occur. The cost savings should be based on efficiency gains, as well as the increased productivity that can be harnessed in other areas of the business by freeing up staff from administration tasks. This ensures companies do not over-invest at the beginning before seeing initial results and if changes are necessary they do not cannibalise potential ROI and companies can still potentially switch to other viable alternative use cases.

Why should companies invest in an AI-powered digital colleague rather than a chatbot?

Before advising companies on what solution they should invest in, it’s important to first establish what they want to achieve. Digital colleagues can provide a far superior level of customer service however, they require greater resource to set up.

Most chatbots are not scalable, once deployed they cannot be integrated into other business areas as they are designed to answer FAQs based on a static set of rules. Unlike digital colleagues, they cannot understand complex questions or perform several tasks at once.

But, the biggest benefit of a cognitive AI solution is that it improves its delivery of tasks over time through machine learning, driving continuous improvement without increased investment. This, and the fact that it is scalable across a wider range of business areas, means it will bring much higher value to a company over the long term.

How can you ensure employees will use the new AI system?

Employee engagement is critical when it comes to digital transformation. We recommend that a new AI deployment is tested within a small group of enthusiastic employees or super users who will give honest feedback on the system’s language and interface to help define the most effective user experience for the business.

What’s great about starting with a group of super users is that if the test is successful, they can help engender a positive outlook towards the new technology amongst their colleagues.

And companies that do not involve employees in the early stages of an AI deployment risk making employees feel like the AI system was forced upon them, preventing widespread adoption amongst the workforce.

How should companies scale up an AI deployment?

My advice to companies looking to a new AI use case is to keep testing it in short 60 to 90-day cycles. It’s important that the solution is continuously developed and perfected before reproducing it elsewhere in the business.

Once the technology is proven we suggest experimenting in two directions. Firstly by scaling the proof of concept – doing more of the same activity at a greater scale with marginal iterations. Secondly by increasing the scope of AI implementation. You can achieve this by experimenting in new areas of the business, with new technologies and brand new initiatives. The second direction is a little more risky but is crucial in becoming a digital enterprise. You should aim to create the right blend of both that fits the culture of your business.

Martin Linstrom, Managing Director, UK&I at IPsoft 

Advertisements

You might also like More from author

Comments are closed.