The Messy Truth of Enterprise AI Strategies
In a recent episode of the Stack Overflow Podcast, host Ryan Donovan interviews Hema Raghavan, co-founder and head of engineering at Kumo.ai, to discuss the complex realities of implementing artificial intelligence in corporate environments. The conversation highlights significant challenges such as pipeline sprawl and the rise of shadow AI, where employees use unapproved AI tools, potentially exposing sensitive company data to external services. Raghavan explains that while executive mandates drive AI adoption across various departments, this often occurs outside IT governance, creating security risks. To address these issues, the discussion explores governance strategies like deploying models within approved platforms and routing requests through monitored gateways. Additionally, Raghavan shares insights into Kumo.ai’s technical approach, which utilizes a single foundation model with on-the-fly database queries to simplify complex feature engineering and prevent broken pipelines. The episode underscores the tension between rapid AI integration and the need for robust security and operational control, offering practical advice for enterprises aiming to leverage AI effectively while maintaining data integrity and regulatory compliance.
Wire timeline
The Messy Truth of Enterprise AI Strategies
In a recent episode of the Stack Overflow Podcast, host Ryan Donovan interviews Hema Raghavan, co-founder and head of engineering at Kumo.ai, to discuss the complex realities of implementing artificial intelligence in corporate environments. The conversation highlights significant challenges such as pipeline sprawl and the rise of shadow AI, where employees use unapproved AI tools, potentially exposing sensitive company data to external services. Raghavan explains that while executive mandates drive AI adoption across various departments, this often occurs outside IT governance, creating security risks. To address these issues, the discussion explores governance strategies like deploying models within approved platforms and routing requests through monitored gateways. Additionally, Raghavan shares insights into Kumo.ai’s technical approach, which utilizes a single foundation model with on-the-fly database queries to simplify complex feature engineering and prevent broken pipelines. The episode underscores the tension between rapid AI integration and the need for robust security and operational control, offering practical advice for enterprises aiming to leverage AI effectively while maintaining data integrity and regulatory compliance.
Stack Overflow Blog