The Inside Scoop

EVOLVE24 NYC, IBM TechXchange 2024, and the Evolving AI Focus

By Dennis Duckworth

Many companies jumped onto the AI bandwagon, buying into the promise of GenAI. However, we’ve read about the significant number of companies with exploratory projects that haven’t made it into production, and some that never will. The exploration is easy—you can find all kinds of fun and interesting things to do with AI across an enterprise. The challenge has been proving the value and return on investment.

The folks at theCUBE Research and ETR just released a study on the current attitude about GenAI and the perceived return on investment. They say that value is being realized, particularly in very basic use cases like text summarization and chatbots. The Harvard Business Review claims that GenAI is becoming a “ground leveler”—it won’t provide a long-lasting leg up on your competitors, but it can help you achieve the same operational efficiencies they’ve achieved or will achieve with AI.

At recent industry events, Cloudera’s own EVOLVE24 New York and IBM’s TechXchange in Las Vegas, a pragmatic attitude towards AI was evident. Companies recognize the need to adopt a level of IT maturity around AI. They need to implement strict governance and security measures for AI just as they have in all the other areas of enterprise data management.

EVOLVE24 New York

One of the big announcements at EVOLVE24 was the expansion of Cloudera’s Enterprise AI Ecosystem. The important word here is “ecosystem”, an acknowledgment that there isn’t one company or tool that provides everything an enterprise needs to adopt AI successfully. Just as other functions within IT (data warehousing, analytics, machine learning, etc.) require different components, AI is no different. Choosing a large language model (LLM) is complex; often, the right approach is using multiple models from multiple sources, not just the biggest model from the loudest vendor

We also launched Cloudera AI Inference, powered by NVIDIA NIM. Inference runs AI in production once the use case has been determined, the ROI has been estimated, and the AI application development and testing have been completed. Cloudera AI Inference helps achieve that ROI and the benefit of the development phase. By embedding NVIDIA technology, we can optimize the performance of AI apps on all processors, especially on NVIDIA GPUs, when running them in production.

IBM TechXchange 2024 (Las Vegas)

IBM’s TechXchange conference reinforced the importance of governance for me. I noticed a recurring theme that companies are getting value from AI, but it’s only sustainable when safeguards are built in. Security and governance are very difficult to build into applications after they have been developed—the same is true for AI applications.

So the AI tide seems to be shifting—use cases and ROI are emerging, but companies must be prudent in adhering to the  IT best practices relied on through other data hype cycles.