Designing for Life (and Death)

How AI transformed my product design practice at Ethos Life

Swarnima Shukla

6 min read

·

Sep 29, 2026

--

Listen

Share

What is Ethos Life? (Bridging the Behavioral Paradox)

Most people assume financial responsibility follows a neat, predictable timeline: finish school, get a job, buy a home, start a family, and secure a financial safety net. But behavioral data reveals a startling disconnect.

According to Ethos’s consumer research study (“When Americans Start Feeling (and Planning) Like Grown-Ups”), Americans report feeling like “real adults” around age 24 typically triggered by moving out or achieving financial independence. Yet, they don’t purchase their first life insurance policy until age 28. This creates a critical four-year “intent-to-action” gap where young families remain completely exposed to financial risk.

Traditional milestones have also decoupled: today, parenthood hits around age 27, while homeownership follows at 29. Nearly 70% of Americans only take financial planning seriously at age 35, usually spurred by health scares or “parenting their aging parents.” Crucially, 50% of Americans state that buying life insurance was the single concrete action that finally made them feel like a true adult.

Ethos Life was built to bridge this four-year inertia gap. The traditional US life insurance industry was historically slow, opaque, and tedious requiring 30 to 60 days of paper forms, intrusive medical exams, and endless phone calls. Ethos transformed this by building a vertically integrated technology platform that delivers personalized policy coverage in minutes.

Scaled Proof of Impact

Solving this deep behavioral friction has driven exceptional platform performance:

• Q2 FY 2026 Revenue: $190 Million, representing +113% Year-over-Year (YoY) growth.
• Cumulative Impact: Surpassed 700,000+ protected families, with 107,847 policies activated in Q2 FY 2026 alone.

Core Product Ecosystem

Rather than relying on static insurance forms, Ethos operates a tightly coupled, three-pillar product ecosystem:

1. Direct-to-Consumer (DTC) Portal: An intuitive digital flow where applicants apply for Term Life, Whole Life, Indexed Universal Life (IUL), and Wills & Estate Planning without medical exams for the vast majority of policies.
2. Agent & Partner Platform: An intelligent operating panel empowering over 10,000 independent agents and FinTech partners with instant quoting, automated identity verification (voice OTP, phone verification), and frictionless e-signatures.
3. Underwriting Workbench (Internal Command Center): The operational platform where human Underwriters step in when automated risk models encounter edge cases, complex medical histories, or high coverage amounts requiring manual evaluation.

Platform Architecture: How the Entities Connect

Ethos functions as a vertically integrated platform connecting consumers, independent agents, automated machine learning decisioning, internal underwriting teams, and top-tier insurance carriers (such as Banner/LGA, Ameritas, TruStage, and Protective).

Press enter or click to view image in full size

The Data Flywheel

As more users interact with the platform, a powerful data flywheel accelerates business growth and risk accuracy:

Press enter or click to view image in full size

The Paradigm Shift: E-Commerce vs. InsurTech (US Audience)

Transitioning from an e-commerce design background in India to a US InsurTech platform was a profound mindset shift. E-commerce design optimizes for impulse, speed, and conversion velocity: countdown timers, prominent primary actions, and frictionless checkouts.

Designing for US life insurance requires the exact opposite approach:

• Impulse vs. Mortality: In e-commerce, friction is the enemy. In life insurance, users must pause and navigate uncomfortable, existential questions: “What would happen to my spouse and children if I died tomorrow?”. Softening or rushing this decision diminishes its emotional weight.
• Radical Clarity over Softening (“Informed Consent”): In consumer products, designers often hide complexity behind simplified UI cards. In US InsurTech, true empathy requires Informed Consent explaining complex medical ratings, debits, and policy terms with absolute transparency so users know exactly what they are choosing.
• Collapsing Emotional Distance in Critiques: Rather than treating screens as abstract pixels, design critiques at Ethos create intentional space to sit with the emotional gravity of mortality. If a flow feels manipulative, fear-driven, or clinical, team members address it immediately.
• US Regulatory & Data Infrastructure: Unlike standard e-commerce checkouts, US InsurTech involves complex identity verification (AML non-documentary checks via GIACT and TransUnion), OFAC sanctions screening, LexisNexis pre-fills, and NACHA-compliant ACH debit authorization guardrails with credit card fallbacks.

Modernizing the Underwriting Engine: Workbench Transformation

The problem space

While Ethos auto-approves the majority of applications through its ML Underwriting Engine, complex cases are referred to human Underwriters in the Referred Underwriting (RUW) bucket.

Historically, Underwriters experienced substantial operational friction:

• Parallel Documentation: Underwriters ran Workbench on one monitor and external Google Docs/Sheets on another, manually copy-pasting applicant profiles, BMI calculations, and Rx summaries. This fragmented workflow added significant friction to every case review.
• Tab Explosion: Over 30% of Workbench cases involve linked policies under the same applicant SSN. Underwriters were opening 10 to 15 separate browser tabs per applicant to cross-reference prior decisions, medical codes, and evidence items.

The Tri-Pillar Deep Dive

To modernize the Workbench, I pioneered an AI-augmented design lifecycle combining human qualitative research with Claude Design and Claude Code.

Press enter or click to view image in full size

Pillar 1: Design Efforts (Human Ownership & Strategic Research)

Instead of delivering surface-level visual refreshes, I initiated a deep problem discovery process:

1. Contextual Inquiry & Shadowing: Conducted structured qualitative interviews and live shadowing sessions with 6+ active Underwriters. I discovered that 7 to 8 out of every 10 cases required exhaustive manual cross-referencing across linked policy tabs.
2. Structuring 6 Core Problem Buckets: Synthesized findings into 6 prioritized operational buckets:

• Bucket 1 (P0): Case Workspace acting as a viewing tool rather than an active working tool.
• Bucket 2 (P0): Clinical evidence review (EHR, Rx, MIB) being manual and unassisted.
• Bucket 3 (P0): Cross-policy navigation requiring tab-switching across linked applications.
• Bucket 4 (P0/P1): Standard Operating Procedures (SOPs) living outside the system in static spreadsheets.
• Bucket 5 (P1/P2): Operational hand-offs across Policy Admin, Salesforce, and Workbench creating manual relay chains.
• Bucket 6 (P0/P1): Case management and assignment bottlenecks.

Engineering Scoping: Collaborated directly with engineering leads to evaluate database query performance, cross-policy table indexing, and API constraints to define realistic MVP scopes.

Pillar 2: AI Intervention (Claude Design & Claude Code Acceleration)

Integrating AI directly into the design workflow fundamentally transformed my speed and thoroughness:

1. Automated UX Audits & Research Synthesis: Fed raw interview transcripts and workshop recordings into Claude to cluster qualitative feedback, surface edge-case themes, and draft structured hypothesis briefs.
2. Design System Integration in Claude Design: Integrated Ethos’s brand tokens, typography, and component specs directly into Claude Design. Instead of generating generic wireframes, Claude produced high-fidelity, interactive HTML prototypes matching our design language on the first pass.
3. Pressure-Testing Unhappy Paths & Edge Cases: AI excels at systematically mapping complex failure modes that designers often overlook during happy-path reviews.

Artifact-to-Code Pipeline: Converted Claude Design’s interactive HTML prototypes directly into clean, modular front-end code using Claude Code. This enabled handing off inspectable, working front-end code directly to developers, eliminating design-to-engineering translation gaps.

Pillar 3: Design Decision Making (Human Judgment & Ergonomics)

While AI provided rapid exploration, human design judgment was essential to ensure usability, brand fidelity, and ergonomics:

1. Stakeholder Alignment Workshops: Co-hosted interactive validation workshops with senior Ops leadership to refine AI-clustered problem buckets against carrier risk.
2. Auditing AI Outputs for UI Standards: Rigorously reviewed every AI-generated layout to ensure strict compliance with spacing tokens, typographic hierarchy, and accessibility standards.
3. The Draggable Notes Decision (Human Ergonomic Fix):

• Feedback: Underwriters needed to take notes while simultaneously reviewing dense medical reports (EHRs, MIB reports, Rx logs).
• AI Initial Proposal: Claude Design initially generated a fixed, sticky side panel that squeezed the main case review layout, forcing horizontal scrolling for dense medical data tables.
• Human Design Decision: I overridden the sticky panel in favor of a floating, draggable notes component. Underwriters could move, dock, or minimize the notes modal anywhere on screen, ensuring they could take structured notes without obstructing critical medical evidence.

Conclusion: The Future of AI-Augmented Design

My experience at Ethos Life highlights a fundamental shift in the product designer’s role. Moving beyond manual pixel tweaking, designers are becoming systems architects and decision curators.

By leveraging Claude Design and Claude Code, I shifted routine exploratory tasks to AI enabling me to map complex edge cases, deliver inspectable front-end code, and focus human energy on qualitative research, ethics, and ergonomic decision-making.

Designing for US InsurTech at Ethos means building software that respects human emotional gravity, empowers Underwriters to make faster, better decisions, and ultimately ensures that over 700,000 families are protected when it matters most.