Industry-Specific Uses of ChatGPT Hold Great Promise

Large language models certainly generate press, yet it is often industry-specific use cases that prove AI’s value to enterprises. That’s why recently announced ChatGPT implementations by Coca-Cola, Fiat, HubSpot, Kia, and Salesforce piqued our interest.   

Harnessing the power of this exciting AI evolution for a specific outcome makes sense. We’ve seen this work in several instances. A specific use with strong training data and model refinement can produce outstanding outcomes. 

For example, Evalueserve’s Insightsfirst Conference Coverage solution has a module trained in technical medical/pharma vocabulary that a layperson or general AI would not understand. This allows the generative AI to produce more accurate conference transcriptions and analysis summaries. It’s 90% accurate, even before our experts review the transcriptions.  

These newer domain-specific approaches to generative text-based chat tools have smart use cases, solid training data sets, and great potential for positive outcomes. Let’s take a deeper look. 

Fiat and Kia

Both auto brands are using ChatGPT as an interface into virtual showrooms. These implementations will benefit from visual pictures and virtual models of the automobile lineups. Perhaps most exciting, though, is that there can be no better source of data about a car than the manufacturer itself.  

Imagine how much easier it will be for buyers to get their questions answered online with this technology, helping potential buyers avoid the car dealership hustle. And it gives both brands more command of their buying experience, ensuring uniformity in how customers receive information.  

Finally, this is only one use case. Kia is already considering moving user manuals to the Internet with a ChatGPT integration to help users find information and troubleshoot problems. 

HubSpot and Salesforce CRM

Two of the largest CRM (customer relationship management) providers in the world, Salesforce and HubSpot, announced GPT implementations within 24 hours of each other. Here are some smart uses of generative AI tailored to very specific uses, helping salespeople craft letters and marketers write targeted campaigns. 

Salesforce has a direct implementation with ChatGPT, and HubSpot has its own flavor of conversational AI that leverages OpenAI tools. Each implementation is unique – Salesforce’s implementation will have a wider spread across more of its products —  but both will use generative AI to craft content targeted toward customers. 

With very specific, defined types of communication, general templated outreach is likely to be somewhat on point. The question is: how will these ChatGPT implementations evolve based on enterprise customer inputs? Will the outputs get more and more customized? Salesforce is offering customers a feature set to “choose their own external model and use natural-language prompts directly within their Salesforce CRM to generate content that continuously adapts to changing customer information and needs in real time.”  

One thing we know: The better the data, the better the output. Reviewing generative AI content for sales and marketing teams will likely be critical to early CRM-focused generative AI successes. 


The Coca-Cola Company intends to use OpenAI’s generative AI tools ChatGPT and DallE, an image generator, to create marketing materialsat speed. It’s a highly specific series of use cases, and given Coca-Cola’s storied marketing prowess, these implementations will likely benefit from great customer and product data. One question might be whether customers will be able to distinguish between communications generated by an AI or a human. We may never know. 


With OpenAI’s ChatGPT and competitor generative AI algorithms available for license, and now an open API into ChatGPT, we will see more domain-specific implementations. We’ve already built several domain-specific implementations with generative AI and are sure to build more as the technology takes off. We expect these implementations to become increasingly relevant and accurate as they are tied to specific use cases and trained with domain-specific data. It should be an exciting year. 


To learn more about Evalueserve’s domain-specific approach to AI, visit our AI Labs webpage. 

Featured image created on Midjourney using this prompt: Virtual showroom for automobiles in abstract, ultramodern, futuristic design style
Geoff Livingston
Vice President Marketing Communications Posts

Latest Posts