AI’s production push puts data platforms on the hook for results

As enterprises rush to incorporate artificial intelligence (AI) into their operations, data platforms are being tasked with delivering measurable business results from these AI pilots. Patience Bogoevici, chief customer officer of Cloudera Inc., highlighted the challenges that hinder the transition from experimentation to production, namely data access, governance and sovereignty issues, and scale.

A survey of over 1,300 IT specialists revealed that data access is the primary barrier, with 80% of respondents citing it as the most significant obstacle. Bogoevici explained that not all data can be moved to the cloud, and some data may need to remain in on-prem or air-gapped environments for security and regulatory reasons. Additionally, a pilot running on a subset of data may not accurately reflect the costs and returns of a full deployment, making it difficult to calculate total cost of ownership and return on investment.

To address these challenges, Cloudera has introduced Anywhere Cloud, a solution set to launch in November, along with a partnership with Mistral AI SAS to incorporate open-weight models into their platform. Bogoevici emphasized the importance of aligning data platform teams with customers' business objectives and ensuring they can fully utilize the technology they have already invested in.

"When I look at things in terms of customers, for me, technology makes sense only when customers are using it," she stated. "I am measured on how customers realize value with Cloudera."