Building Norric’s First AI-Native
Workflow System for Institutional Diligence
Building Norric’s First AI-Native
Workflow System for Institutional Diligence
Norric redesigns how institutional investors operate across the growing public–private market convergence. Investment, IR, and compliance teams still work through scattered documents, repeated DDQs, and manual coordination — slowing decisions and creating unnecessary duplication.
I redesigned Effic’s developer dependent CRM into a no code, marketer driven workflow. In five months, the product grew from a simple A/B tester into a scalable automation system used by 50 plus businesses and driving 6× revenue growth.
TIMELINE
3 months
(2025.10-2025.12)
3 months
(2025.10-2025.12)
ROLE
Foundational Product Designer
Foundational Product Designer
COLLABORATORS
CEO, IR operator, Full-stack developer, Product Designer
CEO, IR operator, Full-stack developer, Product Designer
COMPANY
Norric AI (New York, NY)
Norric AI (New York, NY)
Achievement
-3
-3
-3
-3
5
5
5
hrs
hrs
hrs
hrs
hrs
-3
-3
-3
-3
5
5
5
hrs
hrs
hrs
hrs
hrs
Teams reduced manual searching, copying, and rewriting by consolidating all documents and prior answers into one workflow.
5
5
5
5
X
X
X
X
5
5
5
5
X
X
X
X
Shared states, source trails, and structured workflows eliminated back-and-forth threads and made reviews significantly faster.
6
6
6
6
0
0
0
%
%
%
6
6
6
6
0
0
0
%
%
%
AI recommendations surfaced approved historical responses with transparent sourcing, reducing repeated work and improving consistency.
STarting point
Private-Market diligence isn’t broken because of Insight.
Private-Market diligence isn’t broken because of Insight.
It’s Broken Because of Workflow. Institutional teams don’t lack expertise — they lack structure.
Every DDQ, RFP, and investor request means digging through old folders, rewriting answers, and aligning across departments. The challenge isn’t analysis.
It’s the invisible manual work wrapped around it.
This is the gap Norric set out to solve.
Timeline
Why this problem matters?
Why this problem matters?
What began as a weak A/B testing tool lacked real value. Instead of scaling it, I stopped development, returned to users, and reframed the product around their actual decision-making needs. Within five months, the rebuilt version became the foundation of CRM campaigns for 50+ businesses.
What began as a weak A/B testing tool lacked real value. Instead of scaling it, I stopped development, returned to users, and reframed the product around their actual decision-making needs. Within five months, the rebuilt version became the foundation of CRM campaigns for 50+ businesses.
• 60% of DDQ questions repeat
• Prior answers are scattered across PDFs, desktops, SharePoint, and outdated repositories
• IR teams spend hundreds of hours per quarter searching, copying, and rebuilding content
1st attempt
Our first bet: What if we built a lightweight A/B testing tool for teams like us?
Our first bet: What if we built a lightweight A/B testing tool for teams like us?
First internal A/B testing tool we built in early 2023.
Our first idea was to build a simple A/B tester so teams could optimize message timing.
It worked technically, but it didn’t solve the real problem.
Teams didn’t need better delivery optimization, they needed autonomy from engineers.
Our first idea was to build a simple A/B tester so teams could optimize message timing. It worked technically, but it didn’t solve the real problem. Teams didn’t need better delivery optimization, they needed autonomy from engineers.
The 1st result
Technically complete. Practically irrelevant.
Technically complete. Practically irrelevant.
Our A/B testing tool launched, but adoption stayed flat.
Teams didn’t want more test variants, they wanted a way to run sequences, set timing, and track conversions.
Our team hosted Webinar in mid 2023 to promote effic's 1st launch
Lesson 1 : DTC teams didn't need A/B tests at all.
Lesson 1 : DTC teams didn't need A/B tests at all.
“I just need to send coupons at the right time - not test 5versions of the message.”
E-commerce business owner
“I just need to send coupons at the right time - not test 5versions of the message.”
E-commerce business owner
Lesson 2 : Startup marketers & PM wanted more control.
Lesson 2 : Startup marketers & PM wanted more control.
“A/B testing is great, but I need to run sequences, track conversions, and report results to my CEO.”
Startup marketer
“A/B testing is great, but I need to run sequences, track conversions, and report results to my CEO.”
Startup marketer
Pivot through failure
What were we missing?
What were we missing?
After interviewing 20 plus marketers across different industries, one theme kept repeating:
"I just want to send this message now"
"I just want to send this message now"
Despite different use cases, teams shared the same blocker:
CRM tools required technical steps marketers couldn’t perform.
KEY finding
Small teams didn’t struggle with CRM ideas, they struggled with ownership.
Small teams didn’t struggle with CRM ideas, they struggled with ownership.
Marketers, PMs, and developers each handled a piece of CRM, but no one could run the full process end to end. Every step required engineering, creating delays and breaking momentum.
Marketers, PMs, and developers each handled a piece of CRM, but no one could run the full process end to end. Every step required engineering, creating delays and breaking momentum.
Problem 1
Campaigns sat in developer queues
Marketers had campaigns ready to launch, but every step, from audience segmentation to delivery,
required dev support. This bottleneck delayed launches and killed momentum.
Ask Developer to pull
target audience
⚠️ Bottle neck
⏳ Wait in Dev queue
Marketer/PM has a new
campaign idea
Dev uploads list &
Sets up CRM tool
Marketer writes copy,
but needs Dev to send
Sometimes confirm
details with Sales rep
Ask Developer to pull
target audience
⚠️ Bottle neck
⏳ Wait in Dev queue
Marketer/PM has a new
campaign idea
Dev uploads list &
Sets up CRM tool
Marketer writes copy,
but needs Dev to send
Sometimes confirm
details with Sales rep
Problem 2
Existing CRM tools were built for enterprise scale, not lean teams
Most CRM platforms were built for enterprise scale, complex, expensive, and slow to implement.
Lean teams needed something simpler, faster, and owned by marketers themselves.
One line solution
KEY FINDING
Designing the pivot: from insight to interface
Designing the pivot: from insight to interface
We rebuilt Effic around one principle, marketers should be able to launch campaigns without engineers. The goal was simple: make CRM automation as intuitive as sending an email.
We rebuilt Effic around one principle, marketers should be able to launch campaigns without engineers. The goal was simple: make CRM automation as intuitive as sending an email.
What if CRM automation was as simple as scheduling an email? 📧
What if CRM automation was as simple as scheduling an email? 📧
Started from whiteboard sketches → tested wireframes with users → launched a clean, self-serve UI that marketers could use independently.
UI walkthrough
UI WALKTHROUGH
Jessica, once overwhelmed by CRM,
now handles it effortlessly with Effic.
Jessica, once overwhelmed by CRM,
now handles it effortlessly with Effic.
UI walkthrough
UI WALKTHROUGH
Step 1 : Select Recipients
Who should get this message ?
AS-IS
Export data from Google Sheets, then manually filter to find users who signed up but didn’t purchase within 7 days.
Export data from Google Sheets, then manually filter to find users who signed up but didn’t purchase within 7 days.
TO-BE
In Effic, Jessica selects the “Signup within 7 days” preset and adds the “No purchase” condition — all done in under a minute, no spreadsheet needed.
In Effic, Jessica selects the “Signup within 7 days” preset and adds the “No purchase” condition — all done in under a minute, no spreadsheet needed.
Step 1 : Select Recipients
Who should get this message?
UI walkthrough
Step 2 : Add filters
Who should we avoid sending this to?
AS-IS
Ask a developer to write SQL to exclude irrelevant users (e.g., those who purchased or canceled).
Ask a developer to write SQL to exclude irrelevant users (e.g., those who purchased or canceled).
TO-BE
Jessica applies filters like “Purchase count = 0” or “Signup date within 7 days” with plain-language chips, instantly seeing the user count update live.
Jessica applies filters like “Purchase count = 0” or “Signup date within 7 days” with plain-language chips, instantly seeing the user count update live.
Step 2 : Add filters
Who should we avoid sending this to?
UI WALKTHROUGH
UI walkthrough
Step 3 : Write the message
What do we want to say?
AS-IS
Rewrite messages in Notion or email docs, risking outdated templates and formatting errors.
Rewrite messages in Notion or email docs, risking outdated templates and formatting errors.
TO-BE
Use a built-in editor with smart variables (e.g., {{first_name}}) and preview mode to personalize and verify messages instantly.
Use a built-in editor with smart variables (e.g., {{first_name}}) and preview mode to personalize and verify messages instantly.
UI WALKTHROUGH
Step 3 : Write the message
What do we want to say?
UI walkthrough
Step 4 : Set the timing
When should it go out?
AS-IS
Export user lists to Sendgrid/Braze and schedule separately, often needing dev or PM support.
Export user lists to Sendgrid/Braze and schedule separately, often needing dev or PM support.
TO-BE
Jessica sets delivery rule:
1.Immediate → “Send payment confirmation right after checkout.”
2.Delayed → “3 days after purchase, send a review request.”
3.Time-based → “Send billing reminders every 25th.”
Jessica sets delivery rule:
1.Immediate → “Send payment confirmation right after checkout.”
2.Delayed → “3 days after purchase, send a review request.”
3.Time-based → “Send billing reminders every 25th.”
UI WALKTHROUGH
Step 4 : Set the timing
When should it go out?
UI walkthrough
After launch CRM :
Track performance in real time
AS-IS
Wait days for the data team to pull reports. By the time insights arrived, it was too late to act.
Wait days for the data team to pull reports. By the time insights arrived, it was too late to act.
TO-BE
Jessica checks the Conversion Dashboard to see drop-offs from send → sign-up → purchase in real time. She can also download raw data or tweak metrics instantly—no waiting.
Jessica checks the Conversion Dashboard to see drop-offs from send → sign-up → purchase in real time. She can also download raw data or tweak metrics instantly—no waiting.
UI WALKTHROUGH
After launch CRM
Track performance in real time
Designing effic's ui system
Behind the Scenes: Establishing the Confidence Layer for Complex B2B Tools
Beyond that, as a sole proudct designer, I established Effic’s design language to make hard things feel simple. In the world of clunky B2B tools, our target users wanted something that felt clear, approachable, and genuinely usable, especially for people wearing multiple hats.
Beyond that, as a sole proudct designer, I established Effic’s design language to make hard things feel simple. In the world of clunky B2B tools, our target users wanted something that felt clear, approachable, and genuinely usable, especially for people wearing multiple hats.
To shape Effic’s overall interface, I developed a cohesive design language through key UI components - gradually building a distinct visual system that made the product feel intuitive, approachable, and uniquely ours.
DESIGNING EFFIC'S UI SYSTEM
Behind the scenes : Establishing the confidence layer for complex B2B tools
Testing for real-world use
Validating in the field with our real users.
We ran five test rounds with in-house CRM marketers, using real weekly tasks like re-engagement or segmentation.
We ran five test rounds with in-house CRM marketers, using real weekly tasks like re-engagement or segmentation.
Snapshot: a Project manager using effıc in her workflow
And How did it go? One marketer summed up the impact better than I could:
And How did it go? One marketer summed up the impact better than I could:
“The best part was being able to add or adjust CRM campaigns whenever I needed. Before Effiс, even a small copy change meant waiting on developers. Now I can manage conditions and update messages quickly,
it finally feels like I’m in control.”
“The best part was being able to add or adjust CRM campaigns whenever I needed. Before Effiс, even a small copy change meant waiting on developers. Now I can manage conditions and update messages quickly,
it finally feels like I’m in control.”
Nayul Kim, Product Manager(Online team), Team Sparta
Nayul Kim, Product Manager(Online team), Team Sparta
This confirmed the design goal: putting control back in marketers’ hands and removing the bottlenecks that slowed them down.
This confirmed the design goal: putting control back in marketers’ hands and removing the bottlenecks that slowed them down.
TESTING FOR REAL-WORLD
Validating in the field with our real users
Takeaway
TAKEAWAY
As we conclude, this project taught me…
As we conclude, this project taught me…
0→1 requires speed, but defining
the right problem matters more.
0→1 requires speed,
but defining the right problem matters more.
Early momentum is critical, but I learned that speed without clarity often leads to wasted cycles. Framing the problem precisely made every decision downstream more effective.
Early momentum is critical, but I learned that speed without clarity often leads to wasted cycles. Framing the problem precisely made every decision downstream more effective.
Admit failure early and pivot
using the language of your users.
Admit failure early and pivot using the language of your users.
The turning point came when I stopped defending features and started listening to how marketers themselves described their pain points. Speaking their language unlocked adoption.
The turning point came when I stopped defending features and started listening to how marketers themselves described their pain points. Speaking their language unlocked adoption.
Design is not just about aesthetics, but about creating scalable structures that sustain growth.
Design is not just about aesthetics, but about creating scalable structures that sustain growth.
A polished interface can attract attention, but the real impact comes when the underlying system can flex and scale with the business. That’s where design truly proves its value.
A polished interface can attract attention, but the real impact comes when the underlying system can flex and scale with the business. That’s where design truly proves its value.
Design Challenge
Diligence workflows were scattered across PDFs, emails, and personal folders. Teams rebuilt the same DDQ answers repeatedly with no shared source of truth. This created slow cycles and heavy manual work.
Solution
I designed Norric’s first AI-native workflow system. Documents become structured, searchable, and reusable through Norric AI extraction.
Collaboration, review states, and source trails live in one unified flow.
Impact
DDQs now move through a single centralized system, not 5+ tools.
Repeated manual work dropped significantly with AI-based recommendations. Firms gained clarity, consistency, and a scalable institutional knowledge base.
Designing an AI-native workflow and diligence platform for private-market investors
Reframing tricky CRM,
How I turned a developer-heavy workflow into marketer-driven autonomy
Connect to Content
Add layers or components to make infinite auto-playing slideshows.
I redesigned Effic’s developer dependent CRM into a no code, marketer driven workflow. In five months, the product grew from a simple A/B tester into a scalable automation system used by 50 plus businesses and driving 6× revenue growth.
Connect to Content
Add layers or components to make infinite auto-playing slideshows.
TIMELINE
5 months
(2023-2024)
ROLE
Sole Founding Product Designer (Workflow design, IA, UX, rapid prototyping)
COLLABORATORS
Product Manager, Front/Back-end developer
COMPANY
Teamsparta Inc. (Edtech SMB)
Achievement
5
5
5
5
0
0
0
+
+
+
5
5
5
5
0
0
0
+
+
+
Adopted by SMBs and startups Teams launched campaigns on their own
6
6
6
6
X
X
X
6
6
6
6
X
X
X
Increase in CRM-driven revenue Clients saw measurable lift after switching to Effic
6
6
6
6
0
0
0
%
%
%
6
6
6
6
0
0
0
%
%
%
Marketers launched campaigns without dev handoffs
How it works
Reimagining CRM automation for lean teams
In late 2023, I designed a no-code workflow that lets marketers launch campaigns without engineering bottlenecks. But why we made this? Lean teams couldn’t launch campaigns without engineering help. Even small changes required SQL, data exports, and long handoffs.
So I redesigned Effic into a no code workflow that marketers could run independently.
Timeline
From a stalled MVP to a clear product direction in 5 months
What began as a weak A/B testing tool lacked real value. Instead of scaling it, I stopped development, returned to users, and reframed the product around their actual decision-making needs. Within five months, the rebuilt version became the foundation of CRM campaigns for 50+ businesses.
Start point
Why CRM was the real bottleneck
In edtech, growth depends on retention more than acquisition.
At Team Sparta, thousands of student relationships needed timely messages, reminders, and follow ups. CRM became the core engine for re enrollment and referrals.
Team Sparta is an edtech startup in Korea known for programs like Sparta Coding Club and Tomorrow Learning Camp.By 2023, it had reached $24M in revenue - but that growth depended on managing thousands of student relationships efficiently.
But the tools we used weren’t built for lean teams.
Every update required SQL, data exports, or developer support. Teams moved slowly, experiments stalled, and insights arrived too late.
Impact was slow to measure
Impact was slow to measure
Manual spreadsheets delayed insights and decisions.
Experiments stalled easily
Experiments stalled easily
Even small A/B tests needed developer setup and multiple tools, breaking the creative flow.
Execution dragged
Execution dragged
Campaigns sat in engineering queues instead of reaching users when it mattered most.
Lean teams didn’t need another CRM interface. They needed a faster, clearer way to think and act on user relationships, without relying on engineers.
Lean teams didn’t need another CRM interface. They needed a faster, clearer way to think and act on user relationships, without relying on engineers.
1st attempt
Our first bet: What if we built a lightweight A/B testing tool for teams like us?
First internal A/B testing tool we built in early 2023.
Our first idea was to build a simple A/B tester so teams could optimize message timing. It worked technically, but it didn’t solve the real problem. Teams didn’t need better delivery optimization, they needed autonomy from engineers.
The 1st result
Technically complete. Practically irrelevant.
Our A/B testing tool launched, but adoption stayed flat.
Teams didn’t want more test variants, they wanted a way to run sequences, set timing, and track conversions.
Our A/B testing tool launched, but adoption stayed flat.
Teams didn’t want more test variants, they wanted a way to run sequences, set timing, and track conversions.
Our team hosted Webinar in mid 2023 to promote effic's 1st launch
Lesson 1 : DTC teams didn't need A/B tests at all.
“I just need to send coupons at the right time - not test 5versions of the message.”
E-commerce business owner
“I just need to send coupons at the right time - not test 5versions of the message.”
E-commerce business owner
Lesson 2 : Startup marketers & PM wanted more control.
“A/B testing is great, but I need to run sequences, track conversions, and report results to my CEO.”
Startup marketer
“A/B testing is great, but I need to run sequences, track conversions, and report results to my CEO.”
Startup marketer
Pivot through failure
What were we missing?
After interviewing 20 plus marketers across different industries, one theme kept repeating:
After interviewing 20 plus marketers across different industries, one theme kept repeating:
"I just want to send this message now"
Despite different use cases, teams shared the same blocker:
CRM tools required technical steps marketers couldn’t perform.
KEY finding
Small teams didn’t struggle with CRM ideas, they struggled with ownership.
Marketers, PMs, and developers each handled a piece of CRM, but no one could run the full process end to end. Every step required engineering, creating delays and breaking momentum.
Problem 1
Campaigns sat in developer queues
Marketers had campaigns ready to launch, but every step, from audience segmentation to delivery,
required dev support. This bottleneck delayed launches and killed momentum.
Marketers had campaigns ready to launch, but every step, from audience segmentation to delivery,
required dev support. This bottleneck delayed launches and killed momentum.
Ask Developer to pull
target audience
⚠️ Bottle neck
⏳ Wait in Dev queue
Marketer/PM has a new
campaign idea
Dev uploads list &
Sets up CRM tool
Marketer writes copy,
but needs Dev to send
Sometimes confirm
details with Sales rep
Ask Developer to pull
target audience
⚠️ Bottle neck
⏳ Wait in Dev queue
Marketer/PM has a new
campaign idea
Dev uploads list &
Sets up CRM tool
Marketer writes copy,
but needs Dev to send
Sometimes confirm
details with Sales rep
Problem 2
Existing CRM tools were built for enterprise scale, not lean teams
Most CRM platforms were built for enterprise scale, complex, expensive, and slow to implement.
Lean teams needed something simpler, faster, and owned by marketers themselves.
Most CRM platforms were built for enterprise scale, complex, expensive, and slow to implement.
Lean teams needed something simpler, faster, and owned by marketers themselves.
KEY FINDING
Designing the pivot: from insight to interface
We rebuilt Effic around one principle, marketers should be able to launch campaigns without engineers. The goal was simple: make CRM automation as intuitive as sending an email.
What if CRM automation was as simple as
scheduling an email? 📧
Started from whiteboard sketches → tested wireframes with users → launched a clean, self-serve UI that marketers could use independently.
Started from whiteboard sketches → tested wireframes with users → launched a clean, self-serve UI that marketers could use independently.
UI WALKTHROUGH
Jessica, once overwhelmed by CRM, now handles it effortlessly with Effic.
UI WALKTHROUGH
AS-IS
Export data from Google Sheets, then manually filter to find users who signed up but didn’t purchase within 7 days.
TO-BE
In Effic, Jessica selects the “Signup within 7 days” preset and adds the “No purchase” condition — all done in under a minute, no spreadsheet needed.
Step 1 : Select Recipients
Who should get this message?
AS-IS
Ask a developer to write SQL to exclude irrelevant users (e.g., those who purchased or canceled).
TO-BE
Jessica applies filters like “Purchase count = 0” or “Signup date within 7 days” with plain-language chips, instantly seeing the user count update live.
Step 2 : Add filters
Who should we avoid sending this to?
UI WALKTHROUGH
AS-IS
Rewrite messages in Notion or email docs, risking outdated templates and formatting errors.
TO-BE
Use a built-in editor with smart variables (e.g., {{first_name}}) and preview mode to personalize and verify messages instantly.
UI WALKTHROUGH
Step 3 : Write the message
What do we want to say?
AS-IS
Export user lists to Sendgrid/Braze and schedule separately, often needing dev or PM support.
TO-BE
Jessica sets delivery rule:
1.Immediate → “Send payment confirmation right after checkout.”
2.Delayed → “3 days after purchase, send a review request.”
3.Time-based → “Send billing reminders every 25th.”
UI WALKTHROUGH
Step 4 : Set the timing
When should it go out?
AS-IS
Wait days for the data team to pull reports. By the time insights arrived, it was too late to act.
TO-BE
Jessica checks the Conversion Dashboard to see drop-offs from send → sign-up → purchase in real time. She can also download raw data or tweak metrics instantly—no waiting.
UI WALKTHROUGH
After launch CRM
Track performance in real time
Beyond that, as a sole proudct designer, I established Effic’s design language to make hard things feel simple. In the world of clunky B2B tools, our target users wanted something that felt clear, approachable, and genuinely usable, especially for people wearing multiple hats.
To shape Effic’s overall interface, I developed a cohesive design language through key UI components - gradually building a distinct visual system that made the product feel intuitive, approachable, and uniquely ours.
To shape Effic’s overall interface, I developed a cohesive design language through key UI components - gradually building a distinct visual system that made the product feel intuitive, approachable, and uniquely ours.
DESIGNING EFFIC'S UI SYSTEM
Behind the scenes : Establishing the confidence layer for complex B2B tools
We ran five test rounds with in-house CRM marketers, using real weekly tasks like re-engagement or segmentation.
And How did it go? One marketer summed up the impact better than I could:
“The best part was being able to add or adjust CRM campaigns whenever I needed. Before Effiс, even a small copy change meant waiting on developers. Now I can manage conditions and update messages quickly,
it finally feels like I’m in control.”
Nayul Kim, Product Manager(Online team), Team Sparta
This confirmed the design goal: putting control back in marketers’ hands and removing the bottlenecks that slowed them down.
This confirmed the design goal: putting control back in marketers’ hands and removing the bottlenecks that slowed them down.
TESTING FOR REAL-WORLD
Validating in the field with our real users
TAKEAWAY
As we conclude, this project taught me…
0→1 requires speed, but defining the right problem matters more.
Early momentum is critical, but I learned that speed without clarity often leads to wasted cycles. Framing the problem precisely made every decision downstream more effective.
Admit failure early and pivot using the language of your users.
The turning point came when I stopped defending features and started listening to how marketers themselves described their pain points. Speaking their language unlocked adoption.
Design is not just about aesthetics, but about creating scalable structures that sustain growth.
A polished interface can attract attention, but the real impact comes when the underlying system can flex and scale with the business. That’s where design truly proves its value.
• 60% of DDQ questions repeat
• Prior answers are scattered across PDFs, desktops, SharePoint, and outdated repositories