From Internal Directory to B2B SaaS
From Internal Directory to B2B SaaS
Data Architecture & Monetization Logic for TopTVC
Role
Role
Product Strategy
Information Architecture Workflow
Data QA
Monetization Logic
UX Strategy Research
Tagging System Design
Timeframe
Timeframe
2 years | 0 → 1 Build
Client
Client
Internal Tool → Client-Facing SaaS
Tool
Tool
Adobe XD
Teams
Airtable
Team
Team
1 PM, 3 Engineers
1 UI Designer
8 Content Ops
From Internal Directory to B2B SaaS
Data Architecture & Monetization Logic for TopTVC
Role
UX Strategist ; Tagging System Designer
Timeframe
2 years | 0 → 1 Build
Client
Internal Tool → Client-Facing SaaS
Tool
Adobe XD |Teams | DingDing | Airtable
Teams
Airtable
Team
1 UI Designer , 1 Product manager, 3 Engineers, 8 Content Ops
1 UI Designer
8 Content Ops
Overview
Transformed an internal video library into a monetizable B2B platform.
Designed the data architecture that accelerates creative teams from inspiration to execution.
Inspiration was easy to find, but hard to validate.
High Noise: Public search lacked production reality.
Siloed Data: Best assets were trapped in private folders.
Challenge
Inspiration was easy to find, but hard to validate.
High Noise: Public search lacked production reality.
Siloed Data: Best assets were trapped in private folders.
Challenge
Approach
We built a decision engine, not just a gallery.
Business Signals: Filter by feasibility and rights, not just style.
Structured Truth: Centralized assets into a governed system.
Approach
We built a decision engine, not just a gallery.
Business Signals: Filter by feasibility and rights, not just style.
Structured Truth: Centralized assets into a governed system.
Impact
4h
Saved in shortlist creation
87%
Tagging accuracy
Internal → External
Validated as a product opportunity
Impact
4h
Saved in shortlist creation
87%
Tagging accuracy
Internal → External
Validated as a product opportunity
Every pitch started the same way:
producers buried in late-night meetings, digging through endless decks and chats.
Every pitch started the same way:
producers buried in late-night meetings, digging through endless decks and chats.
“We knew which directors worked for which style, but it all lived in our heads, chats, and scattered folders.”
Every pitch started the same way:
producers buried in late-night meetings, digging through endless decks and chats.
“We knew which directors worked for which style, but it all lived in our heads, chats, and scattered folders.”
6+
6+
Hours
Hours
lost per brief
lost per brief
3-5
3-5
Campaigns
Campaigns
at once
at once
3-5
3-5
Campaigns
at once
0%
0%
Transparency
Transparency
for agency and clients
for agency and clients
0%
0%
Transparency
for agency and clients
Diagnosed through 20+ stakeholder
interviews and a full workflow audit
Diagnosed through 20+ stakeholder
interviews and a full workflow audit
The real problem wasn’t lack of content.
It was the lack of structure and control.
The real problem wasn’t lack of content. It was the lack of structure and control.
01
01
Wasted Investment
Wasted Investment
Wasted Investment
The company poured budget into maintaining a central library, but with near-zero adoption, it became a cost center rather than an asset.
The company poured budget into maintaining a central library, but with near-zero adoption, it became a cost center rather than an asset.
From Senior Producer
From Senior Producer
02
02
Shadow Workflows
Shadow Workflows
Producers bypassed the system for local hard drives. Valuable assets were trapped in "personal silos," inaccessible to the wider team.
Producers bypassed the system for local hard drives. Valuable assets were trapped in "personal silos," inaccessible to the wider team.
From Agency
From Agency
03
03
Misaligned Taxonomy
Misaligned Taxonomy
Misaligned Taxonomy
The system spoke "Database" (Format, Date), but Producers thought in "Vibes" (Mood, Style)—making search functionally broken.
The system spoke "Database" (Format, Date), but Producers thought in "Vibes" (Mood, Style)—making search functionally broken.
From Production Team
From Production Team
We didn’t need more videos. We needed a smarter way to find the right ones.”
We didn’t need more videos. We needed a smarter way to find the right ones.”
Key Needs: alignment, searchable references, speed.
Key Needs: alignment, searchable references, speed.
Drew
Drew
Senior Producer, Creative Agency
Manages 4–5 pitch projects at once, needs fast creative sourcing to win clients.
We didn’t need more videos. We needed a smarter way to find the right ones.”
Key Needs: alignment, searchable references, speed.
Drew
Senior Producer, Creative Agency
Manages 4–5 pitch projects at once, needs fast creative sourcing to win clients.
To solve this, I standardized the repeatable
parts of the workflow first.
To solve this, I standardized the repeatable
parts of the workflow first.
That turned hidden producer knowledge into a shared system the team could actually use.
Front stage: discovery and proposal building.
Back stage: coordination, resources, and review.
The platform had multiple user types: producers, content ops, agency clients, and internal editors.
Producers played the critical role in shaping briefs, searching for talent, and aligning with clients.
To solve this, I standardized the repeatable
parts of the workflow first.
That turned hidden producer knowledge into a shared system the team could actually use.
Front stage: discovery and proposal building.
Back stage: coordination, resources, and review.
01
01
Standardized Inputs
Standardized Inputs
Standardized Inputs
Replaced vague emails with structured data fields to capture intent early.
02
02
System Initial Matching
System Initial Matching
System Initial Matching
Mapped creative tags to business needs (Budget, Rights) for faster filtering.
03
03
Quality Governance
Quality Governance
Quality Governance
Implemented a review loop to ensure every asset met brand standards.
04
04
Client-Facing Output
Client-Facing Output
Client-Facing Output
Automated the mood board deck, reducing manual formatting time by 90%.
Producer Workflow → New System
Card Sorting with Agency Creatives
Card Sorting with Agency Creatives
Check Tag Dimension Table
Card Sorting with Agency Creatives
Check Tag Dimension Table
The Market Gap, existing tools offered passive storage,
but lacked the context to drive decisions.
The Market Gap, existing tools offered passive storage,
but lacked the context to drive decisions.
We needed a dynamic engine that filtered
by commercial constraints like budget and usage rights.
To guide our search system strategy, I started by analyzing tools our producers were already using including SourceCreative, a widely adopted paid platform used by global agencies. While it offered a massive video archive of directors, its flat structure lacked filters for tone, execution, or campaign relevance. Producers still relied on memory and personal intuition to assess creative fit, which made it hard to scale or delegate.These insights helped clarify our opportunity: build a system that supports decision-making, not just discovery.
Loading...
Monetization Model
To guide our search system strategy, I started by analyzing tools our producers were already using including SourceCreative, a widely adopted paid platform used by global agencies. While it offered a massive video archive of directors, its flat structure lacked filters for tone, execution, or campaign relevance. Producers still relied on memory and personal intuition to assess creative fit, which made it hard to scale or delegate.These insights helped clarify our opportunity: build a system that supports decision-making, not just discovery.
Storage Model Monetizes storage space, not content utility.
Storage Model Monetizes storage space, not content utility.
Access Model Monetizes static viewing rights, lacking discovery logic.
Access Model Monetizes static viewing rights, lacking discovery logic.
TOPTVC
TOPTVC
TOPTVC
Decision Model Monetizes the "Find"
Decision Model Monetizes the "Find"
The Real Challenge
The Real Challenge
Producer knowledge was not shared.
Operations lived across fragmented workflows.
No consistent system connected proposal building with proposal execution.
Producer knowledge was not shared.
Operations lived across fragmented workflows.
No consistent system connected proposal building with proposal execution.
Reframing the Constraint
Reframing the Constraint
Reframing the Constraint
The “Shared Logic”
The “Shared Logic”
Created a visual SOP framework to turn tacit producer knowledge into shared decision support.
Created a visual SOP framework to turn tacit producer knowledge into shared decision support.
Logic over Tech
Logic over Tech
Used clear rules and structured tagging to reduce reliance on individual memory and make discovery more consistent.
Used clear rules and structured tagging to reduce reliance on individual memory and make discovery more consistent.
The “Value” Lock
The “Value” Lock
Brought critical business data into the platform so proposal decisions stayed connected to execution.
Brought critical business data into the platform so proposal decisions stayed connected to execution.
The “Shared Logic”
Studied behavior models, iOS HIG, and accessibility guidelines
Logic over Tech
Focused on micro-interactions that feel emotionally rewarding
The “Value” Lock
Brought critical business data into the platform so proposal decisions stayed connected to execution.
Constraints didn't kill the project. They made it scalable.
a system that works for interns, works for everyone.
Constraints didn't kill the project. They made it scalable.
a system that works for interns, works for everyone.
Constraints didn't kill the project. They made it scalable.
a system that works for interns, works for everyone.
Internal Validation →
External Product
But beyond internal use.
I recognized a broader opportunity.
Our internal metrics proved the engine worked
This efficiency wasn't just an operational win; it was a marketable asset.
Our internal metrics proved the engine worked
This efficiency wasn't just an operational win; it was a marketable asset.
Metric
Metric
Time to create a viable shortlist
Time to create a viable shortlist
Tagging Accuracy
Tagging Accuracy
Internal Tag Rework
Internal Tag Rework
Before
Before
6+ hours
6+ hours
68%
68%
26%
26%
After |
|---|
After
⬇︎ 2.5 hours
⬇︎ 2.5 hours
⬆︎ 87%
⬆︎ 87%
⬇︎ 9%
⬇︎ 9%
We didn’t just organize tags,
We monetized access
We didn’t just organize tags.
We monetized access.
Loading...
Why This Worked
Why This Worked
Users found value before paying.
Directors got more exposure.
Paid filters matched real creative needs.
Built a new revenue model based on search utility. Reduced creative alignment time from 6 hours to under 2 hours. Teams could filter like I did, but at scale.
Information Architecture of Advanced Search Filters
Information Architecture of Advanced Search Filters
Visualizing the Engine with Generative AI
Leveraging AI tools to reconstruct confidential system logic,
moving beyond static wireframes to high-fidelity conceptualization.
The original backend is confidential. This interface was synthesized using Midjourney/AI to demonstrate the IA logic.
The original backend is confidential. This interface was synthesized using Midjourney/AI to demonstrate the IA logic.
Internal Validation →
External Product
Our internal metrics proved the engine worked
This efficiency wasn't just an operational win; it was a marketable asset.
Metric
Time to create a viable shortlist
Tagging Accuracy
Internal Tag Rework
Before
6+ hours
68%
26%
After |
|---|
⬇︎ 2.5 hours
⬆︎ 87%
⬇︎ 9%
We didn’t just organize tags,
We monetized access
Loading...
Why This Worked
Users found value before paying.
Directors got more exposure.
Paid filters matched real creative needs.
Information Architecture of Advanced Search Filters
Visualizing the Engine with Generative AI
Leveraging AI tools to reconstruct confidential system logic,
moving beyond static wireframes to high-fidelity conceptualization.
The original backend is confidential. This interface was synthesized using Midjourney/AI to demonstrate the IA logic.
To ensure reliability,
I added QA review loops and random audits.
To ensure reliability,
I added QA review loops and random audits.
High-quality data is our product. I engineered a 'Human-in-the-loop' review system to prevent data rot and ensure client trust."
We ran 3+ rounds of usability testing with content managers and editors.
Due to NDA constraints, this flowchart is a reconstructed visualization of the actual upload-review system I designed
Due to NDA constraints, this flowchart is a reconstructed visualization of the actual upload-review system I designed
To ensure reliability,
I added QA review loops and random audits.
High-quality data is our product. I engineered a 'Human-in-the-loop' review system to prevent data rot and ensure client trust."
We ran 3+ rounds of usability testing with content managers and editors.
Due to NDA constraints, this flowchart is a reconstructed visualization
of the actual upload-review system I designed
The 2026 Vision: From Tool to Ecosystem
The 2026 Vision: From Tool to Ecosystem
Find
Monetize the "Find" : Use Multimodal AI to find the perfect shot in seconds, skipping the keywords.
Find
Monetize the "Find" : Use Multimodal AI to find the perfect shot in seconds, skipping the keywords.
Plan
Generate the Pitch Don't just collect videos, draging them into a canvas to auto-generate storyboards.
Plan
Generate the Pitch Don't just collect videos, draging them into a canvas to auto-generate storyboards.
Collaborate
Own the Workflow Stop emailing files. Share work and get client feedback directly inside the platform.
Collaborate
Own the Workflow Stop emailing files. Share work and get client feedback directly inside the platform.
