In this episode, Amir Bormand sits down with John Cottongim, Co-Founder and CTO of Roots Automation, to explore how automation and generative AI (GenAI) are revolutionizing the insurance industry. John brings deep expertise from his 13-year tenure at AIG, sharing insights on legacy systems, data challenges, and the evolving role of digital co-workers. He discusses the intersection of process automation, data quality, AI adoption, and how companies can strategically implement AI without disrupting their workflows.
Key Takeaways
🔹 Insurance’s Tech Debt is Massive – Large insurance companies operate on hundreds or even thousands of legacy systems, making modernization complex.
🔹 Process Automation is a Game-Changer – Digital co-workers can handle high-volume, structured processes like claims intake and underwriting, improving efficiency without drastic process overhauls.
🔹 The Data Quality Dilemma – Poorly structured, inconsistent data has been a barrier to AI adoption, leading many insurance firms to struggle with leveraging data lakes effectively.
🔹 Choosing What to Automate – The best automation candidates are high-volume, repetitive tasks with clear guidelines. Avoid low-impact projects that don’t drive ROI.
🔹 The Future of AI in Insurance – Instead of a single "super AI," the industry will see a network of narrow AI models trained specifically for underwriting, claims processing, and other specialized tasks.
🔹 Adoption Challenges – Employees are often comfortable with inefficient processes. Successful automation integrates smoothly without demanding drastic workflow changes.
Timestamped Highlights
⏳ [00:01:00] – What is Roots Automation? Introducing InsureGPT, digital co-workers designed for insurance automation.
⏳ [00:03:00] – The massive legacy tech burden in insurance: Why do companies still run on decades-old systems?
⏳ [00:06:30] – The data problem in insurance: Why data lakes failed and how AI can improve data capture and structure.
⏳ [00:10:00] – How to pick the right automation projects: Find high-impact, high-volume tasks rather than niche, low-risk ones.
⏳ [00:15:00] – The tech debt of insurance processes: How decades of quick fixes have created a tangled mess of workflows.
⏳ [00:19:00] – Change management and adoption: How to help teams transition from manual work to digital co-workers.
⏳ [00:24:00] – The rise of narrow AI: Why insurance AI models must be specialized rather than all-purpose.
⏳ [00:26:00] – The future of AI in insurance: How companies will need to orchestrate multiple AI models for efficiency.
Quote of the Episode
"The biggest mistake in automation is choosing a process that’s ‘safe’ but not impactful. If it’s not a high-value problem, no one will care when you solve it." – John Cottongim
Connect with John Cottongim
📩 Email: john@rootsautomation.com
🌐 Company Website: Roots Automation
Enjoying the Podcast?
💡 Share this episode with a colleague who’s interested in AI and automation in insurance!📩 Subscribe, rate, and leave a review to help us keep delivering great conversations.