Big Ideas in App Architecture · Cockroach Labs

Modernizing Insurance Application Architecture at New York Life

·44 min·4 clips

Mike Murphy explains why a life insurance policy from 1980 can't be changed, even if the software evolves.

Mike Murphy, Corporate Vice President at New York Life, joins the host to discuss modernizing application architecture in the life insurance industry. Murphy has decades of experience building and implementing policy administration software for insurers. The episode explores the unique technological challenges of an industry built on long-term contracts and legacy systems. Life insurance involves diverse products like term life, whole life, and variable universal life, each with distinct computational and data needs. Every policy is a legal contract that cannot be altered, forcing systems to maintain original terms for decades. Carriers manage legacy products that may have sold only a few thousand policies but still require ongoing software support. Policies include complex elements like rate tables, state-specific rules, commission hierarchies for agents, and optional add-ons called riders. Monthly processing involves payments, withdrawals, fund rebalancing, commission calculations, and generating correspondence. Insurers must maintain detailed historical records, sometimes for up to 75 years, though legal requirements are shorter. A handful of vendors, including Accenture Life Insurance Platform and Oracle, provide the core policy administration systems, with very few new platform sales occurring each year. These systems are essentially low-code/no-code platforms, allowing business users to configure products and rules without hardcoded software changes. Migrating to a new platform is a multi-year endeavor, often taking three to ten years, with some legacy policies simply "run off" instead of being converted. A major technical challenge is processing millions of policy transactions in a narrow nightly window to meet banking deadlines for fund movements. Quarter-end processing requires evaluating all policies to calculate reserves and ensure regulatory compliance, creating massive batch workloads. Horizontal scaling is critical for handling these batch peaks, with cost ideally being net-neutral compared to slower vertical scaling. Using sequential IDs for database keys creates hotspots, as the most active policies are often the newest; GUIDs help distribute the load. Effective horizontal scaling requires designing software so that all records for a single policy share a common key, keeping input/output on one node and minimizing cross-node chatter. The primary benefit of faster processing is increased resilience, providing more time to recover from failures before hard external deadlines. Beyond batch systems, companies also invest in customer and agent-facing applications, with strategies varying between agent-assisted and direct-to-consumer models. Murphy sees AI potentially accelerating legacy system migrations by reverse-engineering old code and converting it to modern rules. He also mentions AI applications in areas like claims processing, fraud detection, and product development through data analysis. The conversation is technical and educational, focusing on architectural trade-offs and industry-specific constraints. It maintains a conversational tone, blending historical context with forward-looking solutions. This episode is ideal for enterprise architects, software engineers in regulated industries, and anyone curious about modernizing complex legacy systems. Listeners seeking a high-level business strategy discussion or consumer-focused tech trends might find it too granular.

As heard by us

Insurance modernization feels concrete when policy systems, payment timing, and operational discipline all have to line up.

The episode keeps coming back to policy administration, mainframe-era habits, and the plain work of keeping data and money lined up. Mike Murphy uses the New York Life challenge to show why older platforms last so long, and what has to shift before a move forward is possible.

Read the full review in PlayNext →

Why you'd press play

You get a grounded look at what it takes to modernize a legacy policy administration system.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Big Ideas in App Architecture