Blackbaud | GrantsConnect
I led a cross-functional team to transform GrantsConnect from a service-dependent, 200+ day onboarding process into a modular, AI-powered SaaS platform. Our approach—object-oriented UX, modular IA, and 48+ stakeholder interviews- drove rapid, measurable results.
Role
Role
Lead UX Strategist & IA Architect
Lead UX Strategist & IA Architect
Year
Year
2025
2025
Type
Type
Product Design
Product Design
Techstack
Techstack
Figma, Claude, Vercel, ChadeCN
Figma, Claude, Vercel, ChadeCN
↓85%
Onboarding time
↓80%
Support needs
+7.5
User NPS
↓85%
Onboarding time
↓80%
Support needs
+7.5
User NPS
↓85%
Onboarding time
↓80%
Support needs
+7.5
User NPS
Background
Why are we here?
GrantsConnect was built for Fortune 500 CSR teams and nonprofits to manage grantmaking at scale. Leadership brought me in after identifying onboarding as the key blocker to growth: setup took over 200 days and required heavy manual support, threatening enterprise expansion. We needed a radical shift in approach to solve this challenge.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Background
Why are we here?
GrantsConnect was built for Fortune 500 CSR teams and nonprofits to manage grantmaking at scale. Leadership brought me in after identifying onboarding as the key blocker to growth: setup took over 200 days and required heavy manual support, threatening enterprise expansion. We needed a radical shift in approach to solve this challenge.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Background
Why are we here?
GrantsConnect was built for Fortune 500 CSR teams and nonprofits to manage grantmaking at scale. Leadership brought me in after identifying onboarding as the key blocker to growth: setup took over 200 days and required heavy manual support, threatening enterprise expansion. We needed a radical shift in approach to solve this challenge.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Challenge
What are we solving?
Despite high demand, onboarding was slow and manual—over 200 days per customer, with just one client ever self-configuring. The root problem was structural: fragmented IA, deep modal flows, and disconnected setup silos.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Challenge
What are we solving?
Despite high demand, onboarding was slow and manual—over 200 days per customer, with just one client ever self-configuring. The root problem was structural: fragmented IA, deep modal flows, and disconnected setup silos.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Challenge
What are we solving?
Despite high demand, onboarding was slow and manual—over 200 days per customer, with just one client ever self-configuring. The root problem was structural: fragmented IA, deep modal flows, and disconnected setup silos.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Research
Is this a problem?
To validate and deeply understand the problem, I:
Conducted a comprehensive systems audit across the full grant lifecycle.
Led 48+ interviews with grant managers, Blackbaud implementation specialists, and grant applicants.
Mapped and analyzed the three most-used grant types, surfacing recurring field duplication, “object drift,” and process bottlenecks.
Shadowed users and implementation specialists to directly observe pain points, including up to four nested modal layers per key task and widespread fear of breaking logic with changes.
Identified contradictions between product marketing (“intuitive”) and actual user experience (complex, fragmented, error-prone).
Produced OOUX object maps and component matrices as concrete artifacts of the platform’s structure and user flow breakdowns.
Scenario Mapping:
I mapped four core grantmaking scenarios—Invite-Only, Open Call, Tiered Approval, and Compliance-Focused—surfacing pain points like hand-offs, compliance silos, and missing outputs that required later reconstruction.
Invite-Only: Fastest, but reporting pain due to informal process.
Open Call: Manual due diligence as main friction.
Tiered Approval: Bottleneck at multi-level routing, outcome timing uneven by amount.
Compliance-Focused: Process stalls at repeated checks, with missing evidence at the end.
Takeaway: Documentation and training couldn’t solve a fundamentally architectural problem; systemic change was needed.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Research
Is this a problem?
To validate and deeply understand the problem, I:
Conducted a comprehensive systems audit across the full grant lifecycle.
Led 48+ interviews with grant managers, Blackbaud implementation specialists, and grant applicants.
Mapped and analyzed the three most-used grant types, surfacing recurring field duplication, “object drift,” and process bottlenecks.
Shadowed users and implementation specialists to directly observe pain points, including up to four nested modal layers per key task and widespread fear of breaking logic with changes.
Identified contradictions between product marketing (“intuitive”) and actual user experience (complex, fragmented, error-prone).
Produced OOUX object maps and component matrices as concrete artifacts of the platform’s structure and user flow breakdowns.
Scenario Mapping:
I mapped four core grantmaking scenarios—Invite-Only, Open Call, Tiered Approval, and Compliance-Focused—surfacing pain points like hand-offs, compliance silos, and missing outputs that required later reconstruction.
Invite-Only: Fastest, but reporting pain due to informal process.
Open Call: Manual due diligence as main friction.
Tiered Approval: Bottleneck at multi-level routing, outcome timing uneven by amount.
Compliance-Focused: Process stalls at repeated checks, with missing evidence at the end.
Takeaway: Documentation and training couldn’t solve a fundamentally architectural problem; systemic change was needed.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Research
Is this a problem?
To validate and deeply understand the problem, I:
Conducted a comprehensive systems audit across the full grant lifecycle.
Led 48+ interviews with grant managers, Blackbaud implementation specialists, and grant applicants.
Mapped and analyzed the three most-used grant types, surfacing recurring field duplication, “object drift,” and process bottlenecks.
Shadowed users and implementation specialists to directly observe pain points, including up to four nested modal layers per key task and widespread fear of breaking logic with changes.
Identified contradictions between product marketing (“intuitive”) and actual user experience (complex, fragmented, error-prone).
Produced OOUX object maps and component matrices as concrete artifacts of the platform’s structure and user flow breakdowns.
Scenario Mapping:
I mapped four core grantmaking scenarios—Invite-Only, Open Call, Tiered Approval, and Compliance-Focused—surfacing pain points like hand-offs, compliance silos, and missing outputs that required later reconstruction.
Invite-Only: Fastest, but reporting pain due to informal process.
Open Call: Manual due diligence as main friction.
Tiered Approval: Bottleneck at multi-level routing, outcome timing uneven by amount.
Compliance-Focused: Process stalls at repeated checks, with missing evidence at the end.
Takeaway: Documentation and training couldn’t solve a fundamentally architectural problem; systemic change was needed.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Pivot
Should we turn around?
I saw legacy thinking limiting progress, so I initiated a strategic pivot—pausing feature work to invest in foundational user research. This reset the GrantsConnect vision and anchored our solution in real user needs.
I specifically investigated user anxiety about downstream breakages in conditional logic, outdated field copies, and nested budget/workflow modals before designing solutions.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Pivot
Should we turn around?
I saw legacy thinking limiting progress, so I initiated a strategic pivot—pausing feature work to invest in foundational user research. This reset the GrantsConnect vision and anchored our solution in real user needs.
I specifically investigated user anxiety about downstream breakages in conditional logic, outdated field copies, and nested budget/workflow modals before designing solutions.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Pivot
Should we turn around?
I saw legacy thinking limiting progress, so I initiated a strategic pivot—pausing feature work to invest in foundational user research. This reset the GrantsConnect vision and anchored our solution in real user needs.
I specifically investigated user anxiety about downstream breakages in conditional logic, outdated field copies, and nested budget/workflow modals before designing solutions.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Solution
Did we fix it?
Systems Thinking: I conducted a comprehensive system audit and led 48+ interviews with grant managers, implementation specialists, and applicants. By mapping the three most-used grant types and their forms/emails, I identified patterns that enabled a template-first, object-driven approach.
The Challenge: The legacy system’s disconnected features and rigid, multi-screen forms forced grant managers to manually track data, causing friction, errors, and costly delays.
The OOUX Intervention:
To solve this, I introduced the Object-Oriented UX (OOUX) ORCA framework, a methodology for organizing design around core objects, not just features. ORCA stands for Objects, Relationships, Calls-to-Action, and Attributes: a blueprint for a scalable, interconnected data ecosystem.
Applying ORCA, we made foundational entities actionable:
Noun Foraging (Objects): Audited the ecosystems to isolate core objects— Grants, Organizations, Workflows, Requirements,Reviews—(aligning team vocabulary and eliminating ambiguity).
Relationships Mapping: Used a Nested-Object Matrix to reveal deep dependencies. Workflows could dynamically spawn nested Requirements, helping predict database calls and reduce redundant logic.
CTA Mapping (Calls-to-Action): Separated user actions from system triggers, mapping human gates (manager approvals) and AI automations (eligibility checks) to reduce friction.
Attribute Mapping: Broke objects into core content (what users need) and metadata (for sorting, filtering, and logic), enabling modular component design.
Impact: By modeling data structures before any wireframing, we delivered an architecture that supports complex business rules—without UI clutter. This clarified architecture and aligned product, engineering, and implementation teams around a shared language and vision. Workflow creation times dropped, navigation steps were minimized, and engineering handoff accelerated.
AI-Driven Guardrails: I pioneered workflow templates tied to use cases, emails, and forms—giving AI context to guide users and ensure data integrity. Setup included program types, budgets, templates, and company branding, with AI scraping logos and style for scalable, personalized deployments.
User-Centered Change Management: I brought stakeholders in as co-design partners, running workshops and interviews to surface pain points, win advocates, and drive adoption.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Solution
Did we fix it?
Systems Thinking: I conducted a comprehensive system audit and led 48+ interviews with grant managers, implementation specialists, and applicants. By mapping the three most-used grant types and their forms/emails, I identified patterns that enabled a template-first, object-driven approach.
The Challenge: The legacy system’s disconnected features and rigid, multi-screen forms forced grant managers to manually track data, causing friction, errors, and costly delays.
The OOUX Intervention:
To solve this, I introduced the Object-Oriented UX (OOUX) ORCA framework, a methodology for organizing design around core objects, not just features. ORCA stands for Objects, Relationships, Calls-to-Action, and Attributes: a blueprint for a scalable, interconnected data ecosystem.
Applying ORCA, we made foundational entities actionable:
Noun Foraging (Objects): Audited the ecosystems to isolate core objects— Grants, Organizations, Workflows, Requirements,Reviews—(aligning team vocabulary and eliminating ambiguity).
Relationships Mapping: Used a Nested-Object Matrix to reveal deep dependencies. Workflows could dynamically spawn nested Requirements, helping predict database calls and reduce redundant logic.
CTA Mapping (Calls-to-Action): Separated user actions from system triggers, mapping human gates (manager approvals) and AI automations (eligibility checks) to reduce friction.
Attribute Mapping: Broke objects into core content (what users need) and metadata (for sorting, filtering, and logic), enabling modular component design.
Impact: By modeling data structures before any wireframing, we delivered an architecture that supports complex business rules—without UI clutter. This clarified architecture and aligned product, engineering, and implementation teams around a shared language and vision. Workflow creation times dropped, navigation steps were minimized, and engineering handoff accelerated.
AI-Driven Guardrails: I pioneered workflow templates tied to use cases, emails, and forms—giving AI context to guide users and ensure data integrity. Setup included program types, budgets, templates, and company branding, with AI scraping logos and style for scalable, personalized deployments.
User-Centered Change Management: I brought stakeholders in as co-design partners, running workshops and interviews to surface pain points, win advocates, and drive adoption.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Solution
Did we fix it?
Systems Thinking: I conducted a comprehensive system audit and led 48+ interviews with grant managers, implementation specialists, and applicants. By mapping the three most-used grant types and their forms/emails, I identified patterns that enabled a template-first, object-driven approach.
The Challenge: The legacy system’s disconnected features and rigid, multi-screen forms forced grant managers to manually track data, causing friction, errors, and costly delays.
The OOUX Intervention:
To solve this, I introduced the Object-Oriented UX (OOUX) ORCA framework, a methodology for organizing design around core objects, not just features. ORCA stands for Objects, Relationships, Calls-to-Action, and Attributes: a blueprint for a scalable, interconnected data ecosystem.
Applying ORCA, we made foundational entities actionable:
Noun Foraging (Objects): Audited the ecosystems to isolate core objects— Grants, Organizations, Workflows, Requirements,Reviews—(aligning team vocabulary and eliminating ambiguity).
Relationships Mapping: Used a Nested-Object Matrix to reveal deep dependencies. Workflows could dynamically spawn nested Requirements, helping predict database calls and reduce redundant logic.
CTA Mapping (Calls-to-Action): Separated user actions from system triggers, mapping human gates (manager approvals) and AI automations (eligibility checks) to reduce friction.
Attribute Mapping: Broke objects into core content (what users need) and metadata (for sorting, filtering, and logic), enabling modular component design.
Impact: By modeling data structures before any wireframing, we delivered an architecture that supports complex business rules—without UI clutter. This clarified architecture and aligned product, engineering, and implementation teams around a shared language and vision. Workflow creation times dropped, navigation steps were minimized, and engineering handoff accelerated.
AI-Driven Guardrails: I pioneered workflow templates tied to use cases, emails, and forms—giving AI context to guide users and ensure data integrity. Setup included program types, budgets, templates, and company branding, with AI scraping logos and style for scalable, personalized deployments.
User-Centered Change Management: I brought stakeholders in as co-design partners, running workshops and interviews to surface pain points, win advocates, and drive adoption.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Principles
What was our Northstar?
Unified 9 core objects (Applications, Budgets, Programs, etc.) in a single workspace, supported by OOUX and modular IA for scalability and rapid onboarding of new grant types and clients with minimal engineering lift.
Designed the “Builder Trinity” UI (Palette, Canvas, Properties) for user-centered configuration and reduced complexity.
Replaced modal labyrinths with step-by-step, inline setup flows.
Embedded AI nodes (Summarize, Calculate, Translate, Automated Scoring) with confidence thresholds, using Google & Microsoft models stress-tested for compliance and accuracy. Our AI partnerships enabled not just automation, but real adaptability—tailoring onboarding for diverse enterprise needs.
Rolled out error-checking gates, sandbox modes, and strong validation to prevent silent failures.
I prioritized integrated, object-based AI over “AI Studio” silos or band-aid overlays, focusing our efforts only on features that directly supported user autonomy and scalability.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Principles
What was our Northstar?
Unified 9 core objects (Applications, Budgets, Programs, etc.) in a single workspace, supported by OOUX and modular IA for scalability and rapid onboarding of new grant types and clients with minimal engineering lift.
Designed the “Builder Trinity” UI (Palette, Canvas, Properties) for user-centered configuration and reduced complexity.
Replaced modal labyrinths with step-by-step, inline setup flows.
Embedded AI nodes (Summarize, Calculate, Translate, Automated Scoring) with confidence thresholds, using Google & Microsoft models stress-tested for compliance and accuracy. Our AI partnerships enabled not just automation, but real adaptability—tailoring onboarding for diverse enterprise needs.
Rolled out error-checking gates, sandbox modes, and strong validation to prevent silent failures.
I prioritized integrated, object-based AI over “AI Studio” silos or band-aid overlays, focusing our efforts only on features that directly supported user autonomy and scalability.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Principles
What was our Northstar?
Unified 9 core objects (Applications, Budgets, Programs, etc.) in a single workspace, supported by OOUX and modular IA for scalability and rapid onboarding of new grant types and clients with minimal engineering lift.
Designed the “Builder Trinity” UI (Palette, Canvas, Properties) for user-centered configuration and reduced complexity.
Replaced modal labyrinths with step-by-step, inline setup flows.
Embedded AI nodes (Summarize, Calculate, Translate, Automated Scoring) with confidence thresholds, using Google & Microsoft models stress-tested for compliance and accuracy. Our AI partnerships enabled not just automation, but real adaptability—tailoring onboarding for diverse enterprise needs.
Rolled out error-checking gates, sandbox modes, and strong validation to prevent silent failures.
I prioritized integrated, object-based AI over “AI Studio” silos or band-aid overlays, focusing our efforts only on features that directly supported user autonomy and scalability.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Impact
Did we hit something?
Onboarding time: Reduced from 200+ days to under one week.
Self-service adoption: 80% of new implementations require no internal support.
Operational impact: 80% reduction in onboarding resource needs; support tickets dropped sharply.
User sentiment: Net Promoter Score rose to 7.5; user testing showed a 65% drop in time to complete core tasks.
Business impact: SaaS growth accelerated, and GrantsConnect interaction patterns were adopted by other teams.
Recognition: Showcased in bbcon 2025 keynote and referenced by other SaaS teams.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Impact
Did we hit something?
Onboarding time: Reduced from 200+ days to under one week.
Self-service adoption: 80% of new implementations require no internal support.
Operational impact: 80% reduction in onboarding resource needs; support tickets dropped sharply.
User sentiment: Net Promoter Score rose to 7.5; user testing showed a 65% drop in time to complete core tasks.
Business impact: SaaS growth accelerated, and GrantsConnect interaction patterns were adopted by other teams.
Recognition: Showcased in bbcon 2025 keynote and referenced by other SaaS teams.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Impact
Did we hit something?
Onboarding time: Reduced from 200+ days to under one week.
Self-service adoption: 80% of new implementations require no internal support.
Operational impact: 80% reduction in onboarding resource needs; support tickets dropped sharply.
User sentiment: Net Promoter Score rose to 7.5; user testing showed a 65% drop in time to complete core tasks.
Business impact: SaaS growth accelerated, and GrantsConnect interaction patterns were adopted by other teams.
Recognition: Showcased in bbcon 2025 keynote and referenced by other SaaS teams.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Reflection
What did we learn?
Pre-mapping permissions and budgets prevents rework and late-stage churn.
Endless customization requests signal unclear IA; clarity and modularity solve root issues.
Integrating AI at the architecture level, not as an overlay, builds user trust and transparency.
Cross-functional alignment and co-design drive adoption and enthusiasm, not just compliance.
I leveraged Google and Microsoft AI for taxonomy and translation, always validating against live data to ensure compliance and user trust.
Limiting navigation depth and unifying workspaces is far more effective than layering guidance atop complexity.
If I started over, I’d map permission structures and sub-budget relationships earlier and validate by shadowing grant managers’ daily work directly. Early prototypes with real users would have surfaced blockers faster than relying on internal assumptions—combining deep research with modular, AI-powered design.
I continue to apply lessons from this transformation to large-scale, cross-functional design challenges, balancing product strategy, systems thinking, and real-world impact.
AI-Use Transparency
“I used Google and Microsoft’s AI toolkits to scaffold initial taxonomy and automate form translation, but manually validated all object relationships and confidence thresholds against live product behavior, overriding vendor defaults where needed to ensure compliance, accuracy, and user trust.”
I drove this transformation through systems and strategic thinking, rigorous evidence, deep IA and AI expertise, and a commitment to user and business outcomes. GrantsConnect now sets the benchmark for self-service, scalable SaaS in enterprise grant management.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Reflection
What did we learn?
Pre-mapping permissions and budgets prevents rework and late-stage churn.
Endless customization requests signal unclear IA; clarity and modularity solve root issues.
Integrating AI at the architecture level, not as an overlay, builds user trust and transparency.
Cross-functional alignment and co-design drive adoption and enthusiasm, not just compliance.
I leveraged Google and Microsoft AI for taxonomy and translation, always validating against live data to ensure compliance and user trust.
Limiting navigation depth and unifying workspaces is far more effective than layering guidance atop complexity.
If I started over, I’d map permission structures and sub-budget relationships earlier and validate by shadowing grant managers’ daily work directly. Early prototypes with real users would have surfaced blockers faster than relying on internal assumptions—combining deep research with modular, AI-powered design.
I continue to apply lessons from this transformation to large-scale, cross-functional design challenges, balancing product strategy, systems thinking, and real-world impact.
AI-Use Transparency
“I used Google and Microsoft’s AI toolkits to scaffold initial taxonomy and automate form translation, but manually validated all object relationships and confidence thresholds against live product behavior, overriding vendor defaults where needed to ensure compliance, accuracy, and user trust.”
I drove this transformation through systems and strategic thinking, rigorous evidence, deep IA and AI expertise, and a commitment to user and business outcomes. GrantsConnect now sets the benchmark for self-service, scalable SaaS in enterprise grant management.
Enter image caption
Enter image caption
Enter image caption
Enter image caption
Reflection
What did we learn?
Pre-mapping permissions and budgets prevents rework and late-stage churn.
Endless customization requests signal unclear IA; clarity and modularity solve root issues.
Integrating AI at the architecture level, not as an overlay, builds user trust and transparency.
Cross-functional alignment and co-design drive adoption and enthusiasm, not just compliance.
I leveraged Google and Microsoft AI for taxonomy and translation, always validating against live data to ensure compliance and user trust.
Limiting navigation depth and unifying workspaces is far more effective than layering guidance atop complexity.
If I started over, I’d map permission structures and sub-budget relationships earlier and validate by shadowing grant managers’ daily work directly. Early prototypes with real users would have surfaced blockers faster than relying on internal assumptions—combining deep research with modular, AI-powered design.
I continue to apply lessons from this transformation to large-scale, cross-functional design challenges, balancing product strategy, systems thinking, and real-world impact.
AI-Use Transparency
“I used Google and Microsoft’s AI toolkits to scaffold initial taxonomy and automate form translation, but manually validated all object relationships and confidence thresholds against live product behavior, overriding vendor defaults where needed to ensure compliance, accuracy, and user trust.”
I drove this transformation through systems and strategic thinking, rigorous evidence, deep IA and AI expertise, and a commitment to user and business outcomes. GrantsConnect now sets the benchmark for self-service, scalable SaaS in enterprise grant management.
Enter image caption
Enter image caption
Enter image caption
Enter image caption