Boomi CEO Says 95% of Firms Lack 1 Frontier Engineer for Enterprise AI
Updated
Updated · ZDNet · Jul 1
Boomi CEO Says 95% of Firms Lack 1 Frontier Engineer for Enterprise AI
1 articles · Updated · ZDNet · Jul 1
Summary
Steve Lucas said enterprises need a “frontier engineer” — a specialist who understands neural networks and can optimize frontier models for business use — to unlock AI advantage.
Lucas argued the gap is acute: 95% of organizations likely lack even one person who understands how neural networks work, though companies do not need large teams for the role.
He said the job requires enduring skills in data science and neural networking, usually backed by an advanced degree, rather than short-lived AI hiring fads such as prompt, harness or loop engineering.
Lucas estimated fewer than 2,000 to 3,000 people worldwide can build and train models at today’s scale, but said more non-tech enterprises now need experts who can apply that knowledge inside businesses.
He positioned the frontier engineer between a chief AI officer and a forward-deployed engineer, arguing that rare capability will determine which companies extract the most productivity from AI.
What data infrastructure must firms build before a rare frontier engineer can even succeed and deliver real value?
With frontier engineers so rare, how can most companies avoid falling behind the few who can hire them?
The 95% Failure Rate in Enterprise AI: Data-Driven Insights and Strategies for Turning Ambition into Impact
Overview
Recent findings reveal that 95% of enterprise generative AI projects fail to deliver measurable financial impact, highlighting a widespread challenge in turning AI pilots into real business value. This has led to a crucial re-evaluation of AI adoption strategies across industries. The report shows that externally sourced AI tools and partnerships are much more successful—achieving positive financial outcomes 67% of the time—compared to in-house solutions. Vendor-led projects especially outperform in complex fields like medical AI, where deep expertise and ongoing engineering are essential. Despite this, many companies still focus on building their own systems, often missing out on more effective external options.