The MIT study showing 95% of AI projects fail to deliver meaningful ROI isn't about the technology, it's about scoping. Jacob Andra, CEO & Founder of Talbot West, treats AI deployment as an organizational systems problem, not a technology implementation. AI solutions should become force multipliers for a process redesign, not the primary intervention itself.
Jacob also shares his definition of agentic AI: systems making independent decisions as part of larger orchestrations, with quality control agents that must sign off before processes advance.
Topics discussed:
- Why 95% of AI projects fail to deliver ROI due to inadequate scoping of organizational adjacencies and dependencies
- Differentiating AI capabilities across three categories: no human precedent tasks, force multipliers, and autonomous decision-making within orchestrations
- Applying digital transformation methodology that combines business systems optimization expertise with full-spectrum AI and machine learning capabilities
- Identifying revenue opportunities through process redesign before deploying AI solutions as force multipliers for customer-facing teams
- Clarifying agentic AI definitions from digital employees and autonomous systems to decision-makers within orchestrated workflows
- Implementing quality control agents with specific validation criteria that must approve outputs before processes advance to next stages
- Building human-in-the-loop oversight systems where agent managers monitor multiple AI assistants