Editor’s note: Tom Snyder, executive director of rapidly growing Raleigh-based RIoT and a thought leader in the emerging Internet of Things, recently joined WRAL TechWire’s list of top drawer contributors. “Datafication Nation” premiers today. His columns are part of WRAL TechWire’s Startup Monday package.


RALEIGH – You can’t  have Artificial Intelligence (AI) and Machine Learning (ML) systems without large data sets.  It is by applying mathematical and statistical techniques to data that repeatable and predictable patterns can be detected, codified and applied. Generally speaking, the more data that can be brought to bear, the deeper the insights that can be gleaned from it. So how should companies leverage the data they have within their own organizations to extract value?  And what risks exist that you should weigh against the benefits?

I’ll offer a few thoughts here. If you find this interesting, you may be interested to join RIoT’s Developer Day on August 18th at SKEMA on Centennial Campus. This free day of workshops includes experts like Don Shin, CEO of CrossComm and John Petitte, CEO of Amplifi Labs who will each be teaching courses on how to leverage tools like GPT-4 and other AI tools in an enterprise environment.

Creating value

Move Employees to Higher Value Tasks

Look for repeatable processes within an organization that consume a lot of hours of low-level work. For example, customer outreach is something that tends to follow a “more is better” paradigm, but is labor-intensive. By connecting data from Customer Relationship Management (CRM) tools like Salesforce or Hubspot to a capable Large Language Model (LLM) tool like GPT-4, you can automate the generation of high-quality customer communication. This frees employees to spend more time on higher value follow-up, for example, compared to cold-contacting.

Automate Business Processes

Every business needs to make decisions based on a best guess prediction of the future. Leverage the power of statistical AI to improve that decision-making. A good example is to connect databases associated with inventory, procurement and sales. Then use AI for demand forecasting, order planning and overall supply chain efficiency.

Gain Actionable Insights

Fusing marketing campaign data with sales and CRM databases enables AI to bring statistical understanding to what messaging is more or less effective.  Automation can quickly A-B test variations on marketing content, tone and delivery. Empowers your employees to focus on highly effective work that may have been guesswork previously.

Risks to Consider

Accuracy and Compliance

The AI automation can only be as good as the input data.  If you have incomplete or inaccurate data sets, your results will vary.  If a CRM, for example, is capturing email communications with clients and a sales associate has a number of non-business related, side-bar conversations in those emails, an AI may assume those conversation topics make sense for any client and could send mis-guided emails to other customers. And if you are using controlled data, make sure the AI isn’t accidentally breaking regulatory compliance or privacy laws. Even if the “AI did it”, you’ll be held accountable.

Data Leakage

When corporate data is used to train an AI, the future usage of that AI is open to potential reverse engineering by competitors, and the data sets are exposed to possible unauthorized access through poorly secured tools. Consider a bank that uses historical loan data to optimize underwriting decisions. That data invariably will have personally identifiable customer information as well as internal bank financial strategy. As AI is used to automate marketing and new customer outreach activities, it may inadvertently reveal details about that current customer base that were previously trade secrets. With enough output results, a hostile AI may be able to recreate those initial training data sets.

In conclusion, AI and ML systems require large data sets to function.Companies can leverage the data they have within their own organizations to extract value by automating business processes, gaining actionable insights, and moving employees to higher value tasks. However, there are also risks associated with using AI and ML, such as accuracy and compliance issues, and data leakage.

Companies that are considering using AI and ML should carefully weigh the benefits and risks before making a decision. They should also make sure that they have the right data and tools in place to ensure that the AI and ML systems are effective and secure.

If you are interested in learning more about how to leverage AI and ML in your organization, I encourage you to attend RIoT’s Developer Day on August 18th.