Are You Ready to Transform Your Firm with AI?

by HPA Senior Advisor Ed Wiley

There’s this cool, new technology on the block: Artificial Intelligence. Perhaps you’ve heard of it, a little too much?

While “thinking” machines have been around for decades, AI as we know it (or think we know it) has accelerated over the past several years. With Generative AI – via platforms like ChatGPT, DALL-E, Bard, and other new proprietary and open-source alternatives showing up daily – this technology has dominated cocktail conversations and consulting publications! [On a personal aside, my 17-year-old daughter “knows” students who use ChatGPT to write essays, and even the Baby Boomers with whom I work are leaning into AI. Few foresaw the rapidity with which GenAI entered popular culture, embraced by everyday people and businesses alike.]

What is different in this writeup?  We are providing a layman’s guide to the business-technology intersection in AI.  Beyond the hype, AI has already become a game-changer in business, helping companies lead a variety of transformation efforts: streamlining operations and optimizing supply chains to drive efficiency, enhancing data capture and processing to identify patterns and trends that can inform decision-making, personalizing and improving customer experiences through tools like chatbots and virtual assistants, and so on. All these applications have underpinned business growth, particularly for organizations that have asked the right questions about how AI can help transform their business, while laying the foundation for adoption of more emergent technology.

Is your business AI-ready?

Are you thinking about how AI can empower your organization to become more competitive and profitable? Your competitors are certainly wrestling with the same. An astounding 74% of the world’s largest companies surveyed recently by Accenture have already integrated AI as part of their business strategies, even reworking their cloud plans to achieve AI success. Accenture also reported back in 2021 that executives who discussed AI on their earnings calls were 40% more likely to see their firms’ share prices increase.1 Kudos to those perceived as early adopters!

Here are some questions – and a few HPA recommendations – to consider to set your company on the right path:

1. What specific business goals can Artificial Intelligence help your business achieve?

This is a great time to get specific and granular about all the areas where AI can add true value. Embracing AI by treating AI as a business function, ensuring representation in the C-Suite through establishing a CAIO (Chief AI Officer) role, and aligning AI with the business’s strategic objectives, sets the stage for a successful transformation. 

2. Speaking of data, how robust are your data capture protocols and privacy and security measures?

Data quality is the number one factor underlying AI model performance. Training data for both GenAI and traditional (non-Gen) AI must be representative of the context that’s being modeled. Training must include variation similar to what’s expected by the to-be-trained model, and must be great enough in volume to support the sorts of inferences you want to make (for traditional AI) or outcomes you wish to generate (for GenAI). Moreover, companies must maintain compliance with an expanse in data and AI-related regulations (e.g., GDPR; CCPA; EU AI act, and others that are sure to come).

3. Do you have the right infrastructure and staff to support AI investments?

To get a solid ROI on any AI investment, it’s vital to lay a solid foundation. Some Generative AI use cases might require little buildout – the Gen AI first movers (OpenAI/Microsoft, Alphabet, and Nvidia) provide not only APIs for straightforward integration, but also provide general access for individuals to use their applications for personal use at little or no cost. Organizations with rich tech capabilities and infrastructure might choose to build their own GenAI solutions using the growing number of open-source alternatives hitting the marketplace.

4. What about AI talent?

Talent is tricky. Whomever leads an AI initiative should be at a level that provides access to C-level strategy development. This is often a CAIO role, though the best-fitting role will vary from business to business. As I have addressed in my soon-to-be-released book, AI: From Buzzword to Business Function (Shameless plug. Sorry/not sorry!), the CAIO should be someone who understands analytics and AI, but they do not need to be an expert in these areas. More critical than technical expertise is this person’s ability to operate as the interface of AI and business: Not only will the CAIO lead AI strategy for the business, but they must also identify business opportunities for AI and communicate the value of those opportunities to the rest of the executive team.

An early data science team I led for a Fortune 500 company was made up exclusively of machine learning engineers. While this might have been an advantage for a company that has major demands specific to machine learning, I realized quickly that we were short quite a few roles that, in hindsight, would have given more opportunity for business partnership. AI teams need data engineers, software engineers, MLOps and DevOps specialists, and cloud engineers; and businesses must also have folks who can serve as “analytics translators” by serving as the interface between the business, technical specialists and engineers, and people on the operational side.

5. What AI capabilities will match your needs?

You’ve got your business goals, infrastructure, and talent in place, and your data capture and security are golden. Now what? It’s time to figure out what AI technologies will best support your desired outcomes. On the “build or buy” question, will an off-the-shelf offering fit the bill or are you better off building something more customized to your use case?

GenAI offerings have brought a new parallel to the build-or-buy question. Consider Large Language Models (LLMs) like ChatGPT or Bard. Few organizations can afford the multiple millions of dollars required to train their own versions of these massive models from scratch. This doesn’t mean you can’t integrate LLMs for use cases specific to your business. In fact, businesses can partner with a provider (OpenAI and Google currently lead the pack) for use of a foundation LLM (i.e., one trained on a massive, general dataset such as the Internet or the Library of Congress. This model “knows” a great deal in general, but it knows nothing about your specific business. To teach the model about your own business, you conduct additional context training/tuning on proprietary content to tailor that model for your specific use case. This is one “buy” approach.

The “build” approach for GenAI provides an option for more technically adept businesses. “Build” organizations leverage one of the growing numbers of open-source LLM alternatives, and build the training, content generation, and maintenance of that LLM directly into their workflow, affording far greater control of how the model can perform. It’s worth noting this frequently comes with substantially greater complexity for the organization. 

AI has the potential to transform the way businesses conduct most of their activities. That said, adopting AI requires careful consideration and strategic planning. By focusing on these critical questions, companies can evaluate their readiness for AI implementation, mitigate risks, and maximize the upside that AI brings to their business.

ED WILEY is a senior executive and advisor with 25+ years of building, leading, and advising world-class machine learning, AI, and data science teams and projects. He’s worked with companies at various stages, from startup to Fortune 50. He’s also served as a Stanford PhD researcher, Data and SI Senior Advisor with HPA, Consultant with McKinsey, CTO and CDO executive, and academic chair. Ed and his teams have wrestled the thorny issues of using AI across countless contexts, and during his tenure in the field, he’s learned what’s critical in building an AI practice or executing an AI project: In short, he knows what to do – and what NOT to do – when working with AI. 

  1. “The Art of Ai Maturity.” Accenture, 9 May 2023,