Landing the Right Strategy for Your Business as Generative AI Gains Competitive Edge

Presented by Capgemini

Discover the transformative power of generative AI and how it can revolutionize your organization. Join us for the VB Spotlight event, where industry experts will share real-world use cases, tailored solutions, and the secrets to success.

Watch for free on-demand.

AI strategy used to be the domain of CIOs, but with the rise of generative AI, the entire C-suite is now engaged in discussions about its potential. Adoption rates are soaring as executives seek to embrace new use cases and drive innovation. The development and launch of gen AI solutions are also faster and easier than ever before.

“Generative AI shifts us from a data approach to a model approach,” explains Mark Oost, global offer leader, AI, Analytics and Data Science at Capgemini. “With prompt engineering and model fine-tuning, you can showcase the power of these solutions to executives and gain support for new projects.”

Generative AI has become a competitive advantage across industries, and companies must quickly find ways to integrate this technology into their processes and products.

Fast and easy use cases out of the gate

Generative AI has proven effective in two areas: batch-oriented generative AI, which includes content generation such as job descriptions, website and product text, and CRM system information. Real-time generative AI, on the other hand, has gained significant traction in live interactions like chatbots and knowledge search solutions.

“Implementing these use cases is straightforward, especially if you have a wealth of source material within your organization,” says Oost. “End users find the combination of chat and search highly efficient and user-friendly, as it allows for natural conversations.”

Generative AI also enables live personalization using existing data. For example, while shopping online, consumers can request to see products in different contexts, lighting conditions, or even generate a video on the fly.

Security and responsibility in gen AI

Real-time generation poses challenges that require strict guardrails to ensure responsible use. This includes keeping chatbots on message and preventing hate speech or completely fictional responses. To mitigate these risks, it is recommended to move from off-the-shelf models to open-source models designed for specific use cases or industries. Sectors like finance and healthcare, which handle sensitive information, require even stricter restrictions.

Having a company-wide policy on responsible and ethical AI, as well as a comprehensive testing strategy, is crucial.

“When issues arise, people often blame data scientists, but there should always be a human in control, analyzing and testing the model before deployment,” Oost emphasizes. “Given the challenges of identifying issues in generative AI output, A/B testing in experimental environments will be key.”

Scaling beyond the low-hanging fruit

Once a company moves beyond experimentation and tackles more complex challenges, scaling across the organization becomes the primary concern. Cost is often a significant barrier, not in terms of storage, but in the compute costs associated with large models.

“We are entering the era of big models,” says Oost. “Hyperscalers with APIs may not offer sufficient compute power, and scaling may not be affordable. Hosting your own models also requires substantial upfront compute costs.”

As companies shift towards retraining and fine-tuning their own models instead of relying on off-the-shelf solutions, new players will emerge, offering powerful and cost-effective cloud compute services. In the meantime, Oost believes that the investment in compute is worthwhile due to the significant returns that generative AI can deliver.

The real ROI of generative AI

While generative AI may not provide quantifiable cost savings, its true value lies in enhancing production, customer service, and satisfaction. Previously, finding information required hours of searching, but now it is readily available, along with the necessary context to answer strategic questions. Customers increasingly expect flawless and instantaneous interactions, which generative AI can easily provide.

“This is what sets real-time generative AI solutions apart,” explains Oost. “They are fluid, adaptable, and create engaging experiences. They offer frictionless, instant gratification, which leads to significant gains.”

Watch for free on-demand now!

Agenda