AI for Business: Less pAIn, more gAIn!
The news and blog posts (yes, like this one) are filled with hyperbolic claims about Artificial Intelligence (AI) and the future of business. In the last few months alone:
Google has released Gemini Advanced, AgentSpace, and their Willow Quantum Computing chip.
Microsoft has advancements in GPT-4, Copilot, and Azure Cognitive Services, AWS has Bedrock, Titan, SageMaker, and Tranium
I haven’t even mentioned Tesla, Oracle, SalesForce, Meta, or Nvidia’s AI initiatives. It’s nearly impossible to keep up with the latest news. As a CxO, the question becomes how to cut through the noise of all the things you COULD do and identify those things you SHOULD do for your AI strategy.
Highlights:
AI Readiness: Failing to explore AI tools puts your company at risk of falling behind, so the best time to get started is today.
Data Quality Matters: To get business value out of AI, first focus on your data to create a useful library.
Reduce churn and toil via automation and AI
Supercharge your differentiators
The Evolving Landscape of AI
Despite the somewhat exasperated tone of the opening paragraph, do NOT despair! These are indeed exciting times. 2024 saw AI products quickly moving from fantasy to being readily available (and many for free!) via your browser.
Some of the most impressive advancements include AI images and movie clips that appear in seconds based on simple text prompts. TikTok filters and video fakes are amazing (and sometimes scary!) examples of what AI can do. Progress in humanoid robotics is likewise both shocking and thrilling. However, many of the headline-grabbing announcements are not likely to feature tools that will significantly change most industries in the near term.
AI's Business Disruption Power
Where AI is ready to disrupt is through the power of endless computing combined with Large Language Models, Agentic tools, and your data. By now most are aware of the risks of sharing the crown jewels (your company’s data) with commercial tools. But if you’re failing to set up private cloud-enabled models and exploring AI tools to extract additional value from your data… you’re putting your company at risk of falling behind. If that’s the case, the best time to get started is today.
Getting Value from AI with Better Data
You probably already know this, but it bears reinforcement: LLMs are trained on BILLIONS of inputs. The first L is for Large, and they do mean LARGE. It’s an exaggeration to say the latest models represent the collection of all human knowledge, but it’s not all that far off.
That sounds awesome, but my favorite example of why more information is not always a great thing is that if you want help doing your taxes, the book containing your local 2024 tax code is orders of magnitude more useful than the Encyclopedia Brittanica (I checked, it’s still a thing!). Similarly, pointing an LLM at your collections of corporate data can yield highly useful and time-saving results.
3 Questions to Get Your Data AI Ready
Do we have the ability to distinguish between accurate and inaccurate data? (… and who decides?)
Do we leverage metadata to categorize our information?
Is data easily retrievable?
One of the most interesting conversations Burwood had with Higher Education clients this past year was about the evolving role of university librarians. No longer just managing books, University librarians now oversee all data and information created by institutions.
They are key to the DIKW (Data – Information – Knowledge – Wisdom) process, ensuring data quality is crucial for advancing research. Enforcing active data governance within your organization is foundational for AI value; it’s a garbage-in, garbage-out scenario. You need librarians.
Pragmatic recommendations for getting started
Capitalizing on the maturing capabilities of AI is possible TODAY. Here are specific suggestions for getting started:
AI Adoption Baseline: Establish AI Policies and Security
Address FUD (fear, uncertainty, and doubt) by establishing and publishing your corporate policies
Work with your CISO to ensure basic protections are in place
Burwood recently completed a project with a client where they established an “AI Hub” that includes policy documents, training materials, and approved AI tooling. Creating a safe space for testing can lower the chance that your organization will use unapproved tools. This also helps protect your private data.
AI Adoption Opportunity # 1: Identify backlogs - Reduce churn and toil via automation and AI
Look for places in your organization where processing bottlenecks result from backed-up information.
Document AI tools can be used to quickly and inexpensively digitize and take action on paper and form-based data.
AI Agents can be trained to assist with task execution without the rigidity that makes robotic process automation challenging
Creating training Agents based on existing onboarding materials can help to accelerate learning / increase employee productivity in areas where there is high turnover
The Business Value:
Help Desks, Employee Portals, and Customer Websites are great places to start. Burwood has worked with several clients to collect and present information via simple chatbots / intelligent agents. Reductions in call volume and speed to information can be accomplished via small pilot efforts. Positive impacts can be delivered in days or weeks and demonstrate the value of AI, helping establish a business case for larger efforts.
AI Adoption Opportunity # 2: Organize initial experiments - Supercharge your differentiators
Identifying discrete processes in specific areas of your business gives you a great starting point to plan some initial experiments with AI.
Once you’ve determined a specific experiment to run, gathering data (documentation, forms, even relevant communications, etc.) for model training is the next step. Reviewing your data for quality and relevance is key. This is why it is important to start with a well-defined experiment first. Then you aren’t trying to “boil the ocean” with an enterprise-wide data governance effort right from the start.
The results of your AI experiments will help guide you on where (and where NOT) to invest time and resources in further AI development.
Business Value:
Successfully augmenting your existing business processes with AI will set you apart from your competition!
Phone a friend
Burwood has the privilege of working with many leading universities and research institutions across the United States. The opportunity to work with learning organizations means we get to partner on leading-edge solutions. Some pilot efforts fall flat, illustrating the evolving nature of AI. Other efforts yield promising results and real value.
Your organization need not be bleeding edge to gain the insights of our work. Reach out, we’d love to share more about the work we do and how the solutions we’ve built using the latest technologies could benefit you!
