Getty Images/iStockphoto

Guest Post

Generative AI in the enterprise: Lessons learned

The use of generative AI in the enterprise is changing software interactions and customer engagement. Four use cases can help IT leaders better understand the tech's potential.

The use of generative AI is just getting started -- and IT leaders should understand what the technology can and can't do.

There is growing excitement over the development and deployment of generative AI, which is changing how people interact with software and how companies interact with customers. As enterprises adopt this technology for business purposes, there are many lessons to learn along the way.

Amid a flurry of product launches and demonstrations showcasing the capabilities of generative AI and the emergence of new startups, it's essential to pause and consider where the most widespread adoption and initial insights exist as organizations deploy this technology in corporate settings.

4 use cases of generative AI in the enterprise

The number of generative AI use cases is growing. Here are four that IT leaders should understand:

  1. Better customer engagement. For companies that use call centers to tackle customer issues, generative AI -- which includes GPT-3+ and other types of large language models (LLMs) -- can be more accurate than typical conversational chatbots. Generative AI is more adept at understanding tone and reactions, making engagement feel natural for customers. Conversational AI chatbots are among the first use cases companies embraced. These chatbots can access and query internal information and engage with customers in a humanlike manner by responding to inquiries and addressing common issues. Generative AI could improve customer engagement quality, availability and responsiveness for companies with conversational AI, providing an attractive substitute for manually run call centers.
  2. Improved decision-making. Generative AI could potentially make it easier for business users to extract useful insights quickly from large data sets. The technology converts natural language questions into SQL queries run through a company's databases in order to craft an answer that the user can understand. Organizations could potentially gain benefits by improving efficiency through the ability to make better real-time decisions or ask additional questions to find a more thorough answer.
  3. Faster coding. LLMs are precise in various languages, including programming languages. Programmers wanting to create new software code and accompanying documentation could benefit from generative AI by reducing the time to write code by as much as 50%. For example, Microsoft Power Automate, a tool used for robotic process automation, can be programmed via natural language to automate tasks and workflows. This tool reduces the need for large teams of programmers and testers, minimizing the time and effort required to get automation systems up and running.
  4. Optimized content and products. One of the enterprise's primary use cases for generative AI is content and product creation. The technology could help businesses save time and resources by automatically creating new content, such as product descriptions, marketing campaigns and social media posts. Generative AI can also analyze existing product designs and generate new designs based on data. Businesses could potentially create optimized products using the technology based on customer needs and preferences.

With great promise comes challenges

Sanjay Srivastava, chief digital officer, GenPact

The emergence of generative AI poses several challenges for the enterprise.

The relative immaturity of the technology makes it difficult for widespread adoption. Experimenting with new technology takes patience as integration with existing business processes and workflows takes place and as business and IT leaders identify more of the best use cases. The number of best practices and lessons will grow as generative AI becomes more commonplace. Many of these existing issues will become straightforward as the technology matures and application providers integrate generative AI more deeply into their core offerings.

One important problem with generative AI is that it presents information as accurate, even when it's not. The inaccuracies produced by generative AI make the technology unsuitable for permanent application in industries requiring high precision, such as pharmaceutical, medical or financial services. Organizations must meticulously select the most suitable application areas and establish governance and oversight to reduce the risk.

Generative AI requires corporate guidelines regarding data privacy and access to sensitive information, including personal and confidential data. The risk of unintentional intellectual property loss increases when using proprietary data to train publicly accessible LLMs because competitors could have access to the training results. An organization must balance the need for innovation with the associated risks of generative AI by implementing strong policies and carefully designed frameworks.

Strategies for growth and innovation

As more companies share best practices and implement governance policies, generative AI will yield new advancements that harness the technology's possibilities. Companies interested in experimenting can create small test groups to explore generative AI, including how it serves to reconfigure key business processes. Business and IT leaders should work together to evaluate feedback, understand the lessons learned and develop strategies incorporating generative AI.

Generative AI is an evolving technology, and as more use cases become available, every industry must determine the safeguards to adapt to specific industry regulations. Integrating generative AI into everyday enterprise applications means that business processes and workflows will need adjustment.

The enterprise may want to consider exploring generative AI now to capitalize on the technology's potential. Organizations must establish and communicate clear rules and guidelines for the appropriate use and protection of sensitive information. This process can foster innovation while safeguarding the broader interests of the corporation, promoting a smooth, stable transition of generative AI models in the enterprise.

Sanjay Srivastava leads Genpact's digital and technology businesses. He oversees the company's offerings in artificial intelligence, analytics, automation and digital technology services. Before joining Genpact, Sanjay worked as a technology entrepreneur, creating and building four startups from the founding stage to sustainable product businesses. Those firms eventually were individually acquired by Akamai, BMC, FIS and Genpact. Sanjay also held operating leadership roles in other large corporations including Hewlett Packard, Akamai and SunGard (now FIS), where he oversaw product management, global sales, engineering and services businesses.

Next Steps

Conversational AI vs. generative AI: What's the difference?

10 realistic business use cases for ChatGPT

Dig Deeper on IT applications, infrastructure and operations

Cloud Computing
Mobile Computing
Data Center
and ESG