AI Talk: Using AI to help process automation reach peak efficiency

On 10 December 2024, AI Talk host Kevin Craine was joined by Martin Miller, Ex-Director of AI/ML Production Operations, Levi Strauss & Co; Kirsten Jepson, Fractional Product Marketing Leader, Consultant; and Yemi Oluseun, Programme Director - Business Transformation Strategist, The Change Hive.

Views on news 

As organisations negotiate this transformative journey, they will need to overcome challenges in digital transformation to ensure peak efficiency and sustainable growth, such as making the most of their data, transforming their workforce and selecting the right automaton tools. Purpose-built AI solutions may be the answer to how the complexity can be taken out of automation, which are tailored to specific processes. Leveraging real time data can also improve decision making. Ai can impact different parts of an organisation such as marketing and operations differently. However, when talking about AI deployments, RoI must always be part of the conversation.  
 
Reshaping industries with AI

GenAI is still rather new to businesses and at this stage they tend to use it at the back end – especially SMEs. We can also see how companies that use too much AI can lose customers. But there are repetitive tasks where the use of AI is safer, such as sending out reminders and follow-ups to patients. Chat bots can also do a great job as assistants and coaches.

 

Companies that can come up with process flows that make sense prior to automation will be more successful. In other words, companies should get their processes ready before they jump to automation. Some companies that introduced co-pilot at the workplace and found that it had access to sensitive data through HR and other networks actually found that they were losing customers. Where Ai has already made improvements is, for example, customer segmentation and content creation in marketing. AI can be applied to CRMs too to clean the data, to reduce the churn of the data, as well as help you pinpoint the attributes of the data.

 

Although concerns about regulation thwarting innovation in AI is a legitimate concern, biases in AI systems should be pointed out and rectified. AI can actually help with better representing groups that tend to be marginalised too.

The panel’s advice

• AI is still in its infancy. The technology we use most is still machine learning.
• When you see biases and inaccuracies in data, call it out.
• Encourage your staff to self-train in AI.

For Martin’s book, click here.