Details
Category
AI
Enterprise
Fintech
Type
Product Design
Team
5 Designers, Engineers, CEO
Duration
6 months
Norric helps private market teams organize documents, answer DDQs, and collaborate with AI-assisted workflows.
[01]
overview
What is a DDQ?
A Due Diligence Questionnaire is a standardized set of questions that investors send to organizations before making an investment, covering topics such as operations, governance, compliance, and performance.
Institutional teams spend weeks responding to DDQs, RFPs, and compliance requests. Critical information is often buried across spreadsheets, PDFs, and historical submissions, making it difficult to reuse knowledge with confidence.
Our team designed an AI-powered platform that transforms fragmented documents into collaborative, explainable workflows for fundraising and due diligence operations.
[02]
problem
Traditional DDQ workflows are highly manual and risky.
Teams must:
Search through years of historical documents for reusable answers.
Validate information across multiple sources.Coordinate between legal, compliance, and investment teams.
Ensure that outdated responses are not accidentally reused.
Investment teams feared:
Reusing outdated responses.Losing institutional knowledge when employees changed roles.
Lacking visibility into who reviewed or approved content.
Introducing compliance risks through inaccurate information.
In institutional investing, an inaccurate response carries real consequences, from compliance risks to damaged investor relationships.
[03]
design
Design Principles
Design for Trust, Not Just Speed
Users cared more about confidence than automation. Instead of generating answers from scratch, we surfaced:
Match confidence indicators
Supporting evidence
Historical context
Explainable recommendations
Preserve Human Ownership
Financial diligence is inherently collaborative.
Teams needed:
Clear ownership
Review workflows
Approval processes
Accountability across stakeholders
AI became a collaborator—not a decision maker.
Transform Documents into Institutional Memory
The same questions are answered repeatedly across years of fundraising efforts.We designed systems that transformed static documents into reusable organizational knowledge through:
Structured repositories
Historical matching
Shared workspaces
Master DDQ templates
[01]
users
Understanding our Audience
Our primary users were not individual investors, but institutional teams responsible for raising and managing capital, including:
Investment analysts
Investor relations teamsCompliance professionals
Risk managers
ESG specialists
Portfolio managers
Their work required balancing:
Their work required balancing:
These tensions became the foundation of our design decisions.
[02]
solution
Solution
We designed an enterprise platform
centered around three interconnected experiences:
01. Institutional Memory
Rather than storing files as static artifacts, we transformed documents into reusable
organizational knowledge.
Users could:
Upload historical DDQs and reports.
Structure information through projects and repositories.
Reuse previous answers across future submissions.
Maintain a single source of truth.
The goal was not simply better search—but preserving context, ownership, and
institutional expertise over time.
02. Trustworthy AI
Instead of generating answers from scratch, the system prioritized retrieval and
explainability.For every recommendation, users could understand:
Where information originated.
Why documents were matched.
How confident the system was.
What still required manual review.
This approach reinforced a core belief:
Trust is earned through transparency, not automation.
03. Human Governance
AI supported decision-making without replacing it.
We designed collaborative workflows that included:
Assignment systems
Comment threads
Review cycles
Progress tracking
Approval processes
Activity histories
In high-stakes financial environments, users wanted support—not surrender of control.
Key Product Fliws
01. Project Creation & Knowledge Ingestion
Goal: Transform fragmented documents into structured, reusable workspaces.
Users upload DDQs, spreadsheets, reports, and supporting materials while defining
project metadata, asset classes, and ownership structures.
This foundation enables future retrieval, collaboration, and AI-assisted workflows.
Turning unstructured files into organized institutional knowledge. By introducing metadata,
asset classifications, and project contexts upfront, we enabled more trustworthy
downstream AI recommendations.
02. Explainable Question Matching
Goal: Help teams reuse institutional knowledge without compromising accuracy.
Instead of asking AI to generate responses from scratch, we surfaced:
Historical submissions
Supporting evidence
Match confidence indicators
Contextual explanations
Users remained in control of final decisions while understanding exactly why
recommendations appeared.
Designing for confidence rather than convenience. Match explanations, evidence visibility,
and relevance indicators helped users validate information before reuse.
03. Collaborative Review & Ownership
Goal: Preserve accountability throughout complex diligence processes.
DDQs often require multiple stakeholders across legal, compliance, and investment teams.
We designed systems for:
Assignments
Comments
Status tracking
Shared ownership
Review histories
AI became a collaborator—not a decision maker.
Institutional trust depends on clear ownership. By designing explicit review processes, we
ensured accountability remained with people rather than algorithms.
04. Historical Context & Institutional Memory
Goal: Preserve organizational knowledge across years of submissions.
Users could revisit previous decisions, compare historical answers, and understand how
information evolved over time.
Rather than treating each DDQ as a one-off task, we designed a system that continuously
accumulated institutional expertise.
The most valuable knowledge already existed inside organizations—it simply lacked
structure. Historical visibility transformed past work into future leverage.
05. AI-Powered Prioritization & Guidance
Goal: Help teams understand what deserves attention next.
Instead of asking users:
"What would you like me to generate?"
The agent answered questions like:
What requires legal review?
Which responses have low confidence?
What changed since last year?
What information lacks supporting evidence?
What is the highest-risk submission today?
The AI functioned less like a chatbot and more like a strategic partner.
Designing an AI that prioritizes human attention rather than replacing human judgment.
The system helped teams focus on risk, confidence, and next actions instead ofautomation alone.
[01]
design system
Design System
Working from a blank slate, our team established a scalable enterprise design language to
support a rapidly evolving product ecosystem.
Contributions included:
Component libraries
Typography systems
Color foundations
Interaction patterns
Layout structures
Empty, loading, and error states
Design documentation and engineering handoff materials
The system enabled consistency across multiple workflows while supporting highly
information-dense interfaces.
[01]
conclusion
Reflection
Designing Norric was about more than creating an AI-powered diligence platform—it challenged how I approach product design itself.
As an early-stage startup, the product vision was constantly evolving. Requirements shifted, priorities changed, and workflows were frequently redefined. Rather than simply designing screens, our team had to continuously question assumptions, align fragmented ideas, and shape a coherent product experience from the ground up.
These challenges reinforced a core principle of my design process:
By reframing the challenge from "How can AI automate more work?" to "How can AI help people make better, more confident decisions?", we established a product direction centered around:
Explainable AI recommendations
Institutional knowledge reuse
Human oversight and accountability
Collaborative decision-making
Enterprise-scale information architecture
Ultimately, this project taught me that one of a designer's most valuable skills is navigating ambiguity. When product direction isn't fully defined, design becomes a tool for creating clarity—not just polished interfaces.
