Enterprise systems built at Schwab, Bank of America, and State Farm — engineered for capital, risk, and scalable decision systems.
© 2026 Stanford Ashcraft. Portfolio work shown for illustrative purposes.
Design System Walkthroughs
Reducing design drift and compliance riskVega Design System — Foundations
• Shared FinTech design system for advisor platforms
• Defined primitives: color roles, typography scale, spacing, radius
• Button and component variants standardized at the system level
• Token-first approach ensures WCAG 2.0 accessibility by default
• Enables consistent, predictable implementation across desktop and mobileBaselineBefore systemization, UI patterns were fragmented across teams, with inconsistent styling, duplicated components, and manual rework. Designers and engineers relied on ad-hoc decisions, increasing review cycles and regression risk.
FinTech Iconography System
A scalable icon set designed for financial clarity and density. Standardized proportions, stroke weights, and semantic usage to reduce cognitive load and support rapid scanning in data-heavy advisor workflows.
Tokens & Engineering Integration
Design tokens connected Figma components to the Angular codebase, enabling shared definitions for color, spacing, and states. Reduced design-to-dev translation errors while supporting scalable theming and future system expansion.
Component Variants & Dark Mode
Documented component variants for light and dark modes, including tables, inputs, and controls. Ensured contrast compliance, token-driven theming, and consistent behavior across responsive layouts and dual-monitor advisor environments.
Metrics & Impact
Metrics 30 days pre/post adoption (n=428 advisors):
🌟 UI consistency issues ↓ 61%
⏱ Design-to-dev handoff time ↓ 48%
📉 UI regressions ↓ 37%
📈 CSAT for core workflows ↑ 22%
CASE STUDIES
Dashboard OKRs
High-constraint, executive alignment problemKey Outcomes
• Eliminated daily time waste, boosting individual productivity and focus
• Established a single, reliable source for personalized OKR visibility
• Accelerated executive and team alignment in a high-velocity enterprise environmentBaselineAll metrics measured 60 days pre/post launch, n=312 active users
Problem & Opportunity
Engineers and leads wasted 5–12 minutes daily hunting OKRs across disparate tools (Agility, Confluence, scattered dashboards), leading to context-switching friction and delayed alignment. Traditional customizable dashboards suffered from visual noise, slow load times, and stale data.
Research in 30 seconds
What the data tells us
⚠️ Engineers and leads wasted 5–12 min per day hunting
OKRs across Agility, Confluence, and scattered dashboards
Solution & Impact
Prioritized and shipped a lightweight, zero-page-reload modal accessible from Favorites—pulling live data from Agility and auto-personalizing rollups by user ID/role. This delivered instant, relevant insights without overwhelming the core interface.
The Numbers
What the data tells us
🌟 Result → 94 % of users now check OKRs daily (was 31 %)
average session time to view OKRs ↓ 87 % (from ~7 min → <45 s)
System-Level
This shows how Jira data is safely authorized, mapped, normalized, and cached before rendering in Backstage—balancing real-time accuracy with reliability, performance, and clear recovery paths when permissions, schemas, or rate limits fail.
Integrated Quote
Revenue-critical flow under behavioral frictionKey Outcomes
• Significantly reduced abandonment in bundling scenarios
• Enabled faster, more accurate multi-product quotes with dynamic adjustments
• Improved user trust and completion rates through transparent, intuitive progressionBaselineMetrics from A/B test, 60-day period, 180 k+ quote starts
System-Level
This shows how the knowledge system prioritizes deterministic rules first, falls back to search or AI when needed, and continuously improves through feedback—balancing speed, trust, and accuracy for engineers using a shared chat interface.
Gen Hub
Problem & Opportunity
Engineers lost ~4.6 hours per week chasing configs, secrets, runbooks, and scattered Slack threads (validated via internal logs and surveys). 85% of this wasted time stemmed from repeatable, non-coding tasks—creating major productivity drag in a high-velocity environment.
Research in 30 seconds
What the data tells us
⚠️ 85 % of that lost time was on repeatable, non-coding tasks
Solution & Impact
Prioritized and delivered Gen Hub: a unified, intuitive hub with an embedded AI assistant (Sophia) that surfaces relevant knowledge through a single, intelligent search—aggregating disparate sources in real time for instant, contextual results.
The Numbers
What the data tells us
🌟 ↑ 42 % overall satisfaction
↓ 68 % time to onboard a new project (60 days post-launch, n=312)
Productivity recovery under AI and scale pressureKey Outcomes
• Reclaimed ~4.6 hours/week per engineer, directly boosting coding focus and output
• Reduced context-switching on repeatable tasks by targeting 85% of lost time
• Created a scalable foundation for AI-driven knowledge access, accelerating onboarding, debugging, and cross-team collaboration
BaselineAll metrics measured 60 days pre/post launch, n=312 active users
System-Level
This shows how the mobile quote experience orchestrates underwriting, eligibility, and pricing decisions in real time—preserving trust and conversion by offering bundles opportunistically while always protecting a clear auto-only fallback.
Research in 30 seconds
What the data tells us
⚠️ Auto-insurance quote flows had 58–72 % drop-off,
especially when users tried to bundle with renters/home
Problem & Opportunity
Auto-insurance quote flows experienced 58–72% drop-off rates, particularly when users attempted to bundle with renters/home coverage. Long, repetitive forms and disjointed hand-offs across product lines caused frustration, leading users to abandon before viewing personalized pricing.
The Numbers
What the data tells us
🌟 Result → Completion rate ↑ 41 % (from 38 % → 79 %)
average time-to-quote ↓ 82 % (11 min → <2 min)
Solution & Impact
Auto-insurance quote flows experienced 58–72% drop-off rates, particularly when users attempted to bundle with renters/home coverage. Long, repetitive forms and disjointed hand-offs across product lines caused frustration, leading users to abandon before viewing personalized pricing.
DESIGN SYSTEMS
Go to Governace
Go to Governace
Enterprise systems built at Schwab, Bank of America, and State Farm — engineered for capital, risk, and scalable decision systems.
© 2026 Stanford Ashcraft. Portfolio work shown for illustrative purposes.
Dashboard OKRs
High-constraint, executive alignment problem
Key Outcomes
• Eliminated daily time waste, boosting individual productivity and focus
• Established a single, reliable source for personalized OKR visibility
• Accelerated executive and team alignment in a high-velocity enterprise environment
BaselineAll metrics measured 60 days pre/post launch, n=312 active users
System-Level
This shows how Jira data is safely authorized, mapped, normalized, and cached before rendering in Backstage—balancing real-time accuracy with reliability, performance, and clear recovery paths when permissions, schemas, or rate limits fail.
Problem & Opportunity
Engineers and leads wasted 5–12 minutes daily hunting OKRs across disparate tools (Agility, Confluence, scattered dashboards), leading to context-switching friction and delayed alignment. Traditional customizable dashboards suffered from visual noise, slow load times, and stale data.
Research in 30 seconds
What the data tells us
⚠️ Engineers and leads wasted 5–12 min per day hunting
OKRs across Agility, Confluence, and scattered dashboards
Solution & Impact
Prioritized and shipped a lightweight, zero-page-reload modal accessible from Favorites—pulling live data from Agility and auto-personalizing rollups by user ID/role. This delivered instant, relevant insights without overwhelming the core interface.
The Numbers
What the data tells us
🌟 Result → 94 % of users now check OKRs daily (was 31 %)
average session time to view OKRs ↓ 87 % (from ~7 min → <45 s)
Gen Hub
Productivity recovery under AI and scale pressure
Key Outcomes
• Reclaimed ~4.6 hours/week per engineer, directly boosting coding focus and output
• Reduced context-switching on repeatable tasks by targeting 85% of lost time
• Created a scalable foundation for AI-driven knowledge access, accelerating onboarding, debugging, and cross-team collaborationBaselineAll metrics measured 60 days pre/post launch, n=312 active users
System-Level
This shows how the knowledge system prioritizes deterministic rules first, falls back to search or AI when needed, and continuously improves through feedback—balancing speed, trust, and accuracy for engineers using a shared chat interface.
Problem & Opportunity
Engineers lost ~4.6 hours per week chasing configs, secrets, runbooks, and scattered Slack threads (validated via internal logs and surveys). 85% of this wasted time stemmed from repeatable, non-coding tasks—creating major productivity drag in a high-velocity environment.
Research in 30 seconds
What the data tells us
⚠️ 85 % of that lost time was on repeatable, non-coding tasks
Solution & Impact
Prioritized and delivered Gen Hub: a unified, intuitive hub with an embedded AI assistant (Sophia) that surfaces relevant knowledge through a single, intelligent search—aggregating disparate sources in real time for instant, contextual results.
The Numbers
What the data tells us
🌟 ↑ 42 % overall satisfaction
↓ 68 % time to onboard a new project (60 days post-launch, n=312)
Integrated Quote
Revenue-critical flow under behavioral friction
Key Outcomes
• Significantly reduced abandonment in bundling scenarios
• Enabled faster, more accurate multi-product quotes with dynamic adjustments
• Improved user trust and completion rates through transparent, intuitive progressionBaselineMetrics from A/B test, 60-day period, 180 k+ quote starts
System-Level
This shows how the mobile quote experience orchestrates underwriting, eligibility, and pricing decisions in real time—preserving trust and conversion by offering bundles opportunistically while always protecting a clear auto-only fallback.
Problem & Opportunity
Auto-insurance quote flows experienced 58–72% drop-off rates, particularly when users attempted to bundle with renters/home coverage. Long, repetitive forms and disjointed hand-offs across product lines caused frustration, leading users to abandon before viewing personalized pricing.
Research in 30 seconds
What the data tells us
⚠️ Auto-insurance quote flows had 58–72 % drop-off,
especially when users tried to bundle with renters/home
Solution & Impact
Auto-insurance quote flows experienced 58–72% drop-off rates, particularly when users attempted to bundle with renters/home coverage. Long, repetitive forms and disjointed hand-offs across product lines caused frustration, leading users to abandon before viewing personalized pricing.
The Numbers
What the data tells us
🌟 Result → Completion rate ↑ 41 % (from 38 % → 79 %)
average time-to-quote ↓ 82 % (11 min → <2 min)
Design System Walkthroughs
Reducing design drift and compliance risk
Vega Design System — Foundations
• Shared FinTech design system for advisor platforms
• Defined primitives: color roles, typography scale, spacing, radius
• Button and component variants standardized at the system level
• Token-first approach ensures WCAG 2.0 accessibility by default
• Enables consistent, predictable implementation across desktop and mobile
BaselineBefore systemization, UI patterns were fragmented across teams, with inconsistent styling, duplicated components, and manual rework. Designers and engineers relied on ad-hoc decisions, increasing review cycles and regression risk.
FinTech Iconography System
A scalable icon set designed for financial clarity and density. Standardized proportions, stroke weights, and semantic usage to reduce cognitive load and support rapid scanning in data-heavy advisor workflows.
Tokens & Engineering Integration
Design tokens connected Figma components to the Angular codebase, enabling shared definitions for color, spacing, and states. Reduced design-to-dev translation errors while supporting scalable theming and future system expansion.
Component Variants & Dark Mode
Documented component variants for light and dark modes, including tables, inputs, and controls. Ensured contrast compliance, token-driven theming, and consistent behavior across responsive layouts and dual-monitor advisor environments.
Metrics & Impact
Metrics 30 days pre/post adoption (n=428 advisors):
🌟 UI consistency issues ↓ 61%
⏱ Design-to-dev handoff time ↓ 48%
📉 UI regressions ↓ 37%
📈 CSAT for core workflows ↑ 22%
CASE STUDIES
DESIGN SYSTEMS
Research in 30 seconds
What the data tells us
⚠️ Auto-insurance quote flows had 58–72 % drop-off,
especially when users tried to bundle with renters/hom.
The Numbers
What the data tells us
🌟 Result → Completion rate ↑ 41 % (from 38 % → 79 %)
average time-to-quote ↓ 82 % (11 min → <2 min)
© 2026 Stanford Ashcraft. Portfolio work shown for illustrative purposes.
Research in 30 seconds
What the data tells us
⚠️ Auto-insurance quote flows had 58–72 % drop-off,
especially when users tried to bundle with renters/hom.
The Numbers
What the data tells us
🌟 Result → Completion rate ↑ 41 % (from 38 % → 79 %)
average time-to-quote ↓ 82 % (11 min → <2 min)
Go to Governace
Enterprise systems built at Schwab, Bank of America, and State Farm — engineered for capital, risk, and scalable decision systems.
CASE STUDY REVIEW
DESIGN SYSTEMS
Design System Walkthroughs
Reducing design drift and compliance risk
Vega Design System — Foundations
• Shared FinTech design system for advisor platforms
• Defined primitives: color roles, typography scale, spacing, radius
• Button and component variants standardized at the system level
• Token-first approach ensures WCAG 2.0 accessibility by default
• Enables consistent, predictable implementation across desktop and mobileBaselineBefore systemization, UI patterns were fragmented across teams, with inconsistent styling, duplicated components, and manual rework. Designers and engineers relied on ad-hoc decisions, increasing review cycles and regression risk.
FinTech Iconography System
A scalable icon set designed for financial clarity and density. Standardized proportions, stroke weights, and semantic usage to reduce cognitive load and support rapid scanning in data-heavy advisor workflows.
Tokens & Engineering Integration
Design tokens connected Figma components to the Angular codebase, enabling shared definitions for color, spacing, and states. Reduced design-to-dev translation errors while supporting scalable theming and future system expansion.
Component Variants & Dark Mode
Documented component variants for light and dark modes, including tables, inputs, and controls. Ensured contrast compliance, token-driven theming, and consistent behavior across responsive layouts and dual-monitor advisor environments.
Metrics & Impact
Metrics 30 days pre/post adoption (n=428 advisors):
🌟 UI consistency issues ↓ 61%
⏱ Design-to-dev handoff time ↓ 48%
📉 UI regressions ↓ 37%
📈 CSAT for core workflows ↑ 22%
CASE STUDIES
Dashboard OKRs
High-constraint, executive alignment problem
Key Outcomes
• Eliminated daily time waste, boosting individual productivity and focus
• Established a single, reliable source for personalized OKR visibility
• Accelerated executive and team alignment in a high-velocity enterprise environmentBaselineMetrics 30 days pre/post launch, n=428 active users
System-Level
This shows how Jira data is safely authorized, mapped, normalized, and cached before rendering in Backstage—balancing real-time accuracy with reliability, performance, and clear recovery paths when permissions, schemas, or rate limits fail.
System-Level
This shows how the knowledge system prioritizes deterministic rules first, falls back to search or AI when needed, and continuously improves through feedback—balancing speed, trust, and accuracy for engineers using a shared chat interface.
System-Level
This shows how the mobile quote experience orchestrates underwriting, eligibility, and pricing decisions in real time—preserving trust and conversion by offering bundles opportunistically while always protecting a clear auto-only fallback.
Integrated Quote
Revenue-critical flow under behavioral friction
Key Outcomes
• Significantly reduced abandonment in bundling scenarios
• Enabled faster, more accurate multi-product quotes with dynamic adjustments
• Improved user trust and completion rates through transparent, intuitive progressionBaselineMetrics from A/B test, 60-day period, 180 k+ quote starts
Gen Hub
Productivity recovery under AI and scale pressure
Key Outcomes
• Reclaimed ~4.6 hours/week per engineer, directly boosting coding focus and output
• Reduced context-switching on repeatable tasks by targeting 85% of lost time
• Created a scalable foundation for AI-driven knowledge access, accelerating onboarding, debugging, and cross-team collaboration
BaselineAll metrics measured 60 days pre/post launch, n=312 active users
Problem & Opportunity
Engineers and leads wasted 5–12 minutes daily hunting OKRs across disparate tools (Agility, Confluence, scattered dashboards), leading to context-switching friction and delayed alignment. Traditional customizable dashboards suffered from visual noise, slow load times, and stale data.
Solution & Impact
Prioritized and shipped a lightweight, zero-page-reload modal accessible from Favorites—pulling live data from Agility and auto-personalizing rollups by user ID/role. This delivered instant, relevant insights without overwhelming the core interface.
Research in 30 seconds
What the data tells us
⚠️ Engineers and leads wasted 5–12 min per day hunting
OKRs across Agility, Confluence, and scattered dashboards
The Numbers
What the data tells us
🌟 Result → 94 % of users now check OKRs daily (was 31 %)
average session time to view OKRs ↓ 87 % (from ~7 min → <45 s)
Problem & Opportunity
Engineers lost ~4.6 hours per week chasing configs, secrets, runbooks, and scattered Slack threads (validated via internal logs and surveys). 85% of this wasted time stemmed from repeatable, non-coding tasks—creating major productivity drag in a high-velocity environment.
Solution & Impact
Prioritized and delivered Gen Hub: a unified, intuitive hub with an embedded AI assistant (Sophia) that surfaces relevant knowledge through a single, intelligent search—aggregating disparate sources in real time for instant, contextual results.
Research in 30 seconds
What the data tells us
⚠️ 85 % of that lost time was on repeatable, non-coding tasks
The Numbers
What the data tells us
🌟 ↑ 42 % overall satisfaction
↓ 68 % time to onboard a new project (60 days post-launch, n=312)
Problem & Opportunity
Auto-insurance quote flows experienced 58–72% drop-off rates, particularly when users attempted to bundle with renters/home coverage. Long, repetitive forms and disjointed hand-offs across product lines caused frustration, leading users to abandon before viewing personalized pricing.
Solution & Impact
Auto-insurance quote flows experienced 58–72% drop-off rates, particularly when users attempted to bundle with renters/home coverage. Long, repetitive forms and disjointed hand-offs across product lines caused frustration, leading users to abandon before viewing personalized pricing.
Research in 30 seconds
What the data tells us
⚠️ Auto-insurance quote flows had 58–72 % drop-off,
especially when users tried to bundle with renters/hom.
The Numbers
What the data tells us
🌟 Result → Completion rate ↑ 41 % (from 38 % → 79 %)
average time-to-quote ↓ 82 % (11 min → <2 min)
© 2026 Stanford Ashcraft. Portfolio work shown for illustrative purposes.
Go to Governace