Avnet CIO: Navigating the cloud and AI landscape with a practical approach

As CIO of Avnet — one of the largest technology distributors and supply chain solution providers — I’m responsible for the organization’s IT stack and oversee digital transformation and strategy. Two critical areas that underpin our digital approach are cloud and artificial intelligence (AI). 

Cloud and the importance of cost management 

Early in our cloud journey, we learned that costs skyrocket without proper FinOps capabilities and overall governance. But after putting some discipline around it and pinpointing where we can optimize our operations, we have found a better balance.  

That said, we’re not 100% in the cloud. Instead, we focus on use cases that truly elaborate cloud-enabled capabilities. We prioritize those workloads then migrate them to the cloud. However, we’re not looking to move everything to the cloud. Some operations and functions remain internal. 

When we started with generative AI and large language models, we leveraged what providers offered in the cloud. But we knew from the beginning, with our cloud experience and what providers were doing, it was a costly proposition.  

Now that we have a few AI use cases in production, we’re starting to dabble with in-house hosted, managed, small language models or domain-specific language models that don’t need to sit in the cloud. 

What we’re seeing is in line with much of the research, including what IDC has published in relation to the costs about compute, cooling and sustainability. Without proper management, the cloud proposition with AI is going to be very expensive. 

Trying to rationalize how we use AI in conjunction with the cloud is very important. As a distributor, we are a low-margin business, and cloud can be as much as 25% of your operating income. So, we must look at how we deploy AI and cloud in an agile manner. We must be able to react to the business need and be proactive about providing what the business requires. There’s no other way than to embrace it ourselves. 

Beyond productivity: Using AI to help satisfy customers 

When it comes to AI or genAI, just like everyone else, we started with use cases that we can control. These include content generation, sentiment analysis and related areas.  

As we explored these use cases and gained understanding, we started to dabble in other areas. For example, we have an exciting use case for cleaning up our data that leverages genAI as well as non-generative machine learning to help us identify inaccurate product descriptions or incorrect classifications and then clean them up and regenerate accurate, standardized descriptions. 

While this might be driving internal productivity, you also must think of it this way: As a distributor, at any one time, we deal with millions of parts. Our supplier partners keep sending us their price books, spec sheets and product information every quarter. So, having a group of people trying to go through all that data to find inaccuracies is a daunting, almost impossible, task. 

But with AI and genAI capabilities, we can clean up any inaccuracies far more quickly than humans could. Sometimes within as little as 24 hours. That helps us improve our ability to convert and drive business through an improved experience for our customers. 

Another AI example is our design services. We have a design hub that provides customers with self-service capabilities to create bot diagrams and turn them into spec sheets and part lists. With this tool, genAI allows customers to ask questions like, “Help me reduce the physical footprint by 30%” or “How do I drive my go-to-market timeline to three months shorter?” With the right data, the tool can incorporate these requests and create replacement specs, parts lists and suggestions. In this instance, incorporating genAI opens up a new, seamless experience we can offer the customer.

The importance of data and focus 

For organizations looking to incorporate more AI use cases, the right data is critical. That’s why we talk about clean data and AI-ready data. It is so important. Going back to the cloud, we make sure all the data we bring to the cloud allows for downstream consumption. 

Before embarking on any AI initiative, my No. 1 piece of advice is to ensure your data is in the best shape possible. 

The No. 2 thing is to stop being stuck in what I call proof-of-concept (POC) jail. You must prioritize. Look for one, two or three use cases that can scale–and then go scale. Because if you don’t scale, you’re going to be working with hundreds and hundreds of projects and make little progress or impact. 

A recent IDC report found that businesses may be dealing with somewhere around 39 POCs. But then only five go into production. My question is: Why deal with the 39? Just focus on the five and then scale one or two of them. 

Lastly, there are so many providers, and so many models out there. Pick one and try it. Because at the end of the day, you’ll learn from it. If you need more or something different, decide what to do or what it is and go from there. Don’t just try everything. That approach will bring you back to POC jail, handicapping your organization from making genuine progress and impact. 

Max Chan is a skilled, visionary technology executive serving as the CIO at Avnet. He oversees the resources and capabilities of the global IT team and ensures that the organization maintains a robust and optimized IT environment. He also leads the digital system integration business, which delivers digital services and solutions to internal and external customers. Max joined Avnet in 2013 as vice president of IT for Avnet Technology Solutions in Asia Pacific. In 2016, he transferred to the Phoenix corporate headquarters to take on the global application and business relationship management role. Avnet named him CIO in 2019. Before joining Avnet, Max held several IT leadership roles, including CIO, Asia, at VF Corporation, and vice president, IT global supply chain, building efficiency, at Johnson Controls.