CONTEXT
An AI-Powered Sourcing Assistant
Finding suppliers shouldn't feel like detective work. Sourcing Ally helps enterprise teams find qualified suppliers by combining natural language with the accuracy and efficiency of artificial intelligence.
PROJECT DETAILS
My Role
Founding Product Designer
Project Length
3 months
My Team
1 Designer (Me)
1 Project Manager
1 Developer
CEO, CFO, Head of CX, Head of Technology
My Contributions
End-to-End Product Design
AI-Enhanced Workflows
Branding
Marketing Content
SUCCESS METRICS
40% less effort, 90% match accuracy
Supplier discovery went from weeks to just hours, allowing enterprises to make sourcing decisions 3x faster.
What used to take weeks, now takes hours…
…thanks to AI and a smarter way to ask.
An Overwhelming Search for Suppliers
PROBLEM
Enterprises face an overwhelming search for qualified suppliers. Endless databases, outdated information, and unreliable data lead to wasted time, costly setbacks, and missed opportunities.
An AI-Powered Sourcing Assistant
SOLUTION
Sourcing Ally simplifies supplier discovery with a natural language prompt - users describe their sourcing needs in plain language and the system interprets, enhances, and delivers high-quality matches.
Learn about some of the challenges we faced during ideation.
Learn More
Let's dive into how this solution came to life.
Full Case Study | 5-10 min read
BACKGROUND
Product History
Sourcing Ally is the product of Scaleup, a B2B that originated at AWS (Amazon) and went independent in 2023.
Learn more about the product history, solutions that didn't work, and what we learned from them.
Learn More
My Design Approach
Detect
Industry Research
User Insights & Pain Points
Competitive Analysis
Persona Building
Journey Map
Distill
User Flows
Gather UI Inspiration
Define Brand Identity
Design System
UI Design
Design
Prototypes
User Testing
Implement Feedback
Launch
Therefore, enterprises face delays, high costs, and missed opportunities
As a result, enterprise users were stuck in the past - relying on slow, labor-intensive processes that were not only draining their resources and their time, but were preventing them from finding the right suppliers at the right time.
USER INTERVIEWS
AI Usage
Procurement leaders want quality, not quantity
What we learned was that procurement leaders didn't care about quantity but instead wanted quality. They no longer wanted to waste time with suppliers who were unreliable and unfit.
I used ChatGPT to analyze my interview transcripts, cluster common themes, identify pain points, and summarize feedback. As a result, 3 core pain points emerged:
SOURCING TAKES TOO LONG:
Pain Point: On average, it takes 57 days just to go from posting a request to finding a supplier.
Impact: Delays operations, increases costs, and limits business agility.
OUTDATED SEARCH METHODS:
Pain Point: 63% search online and 59% rely on magazines, reports, conferences, and word-of-mouth.
Impact: Inconsistent results, wasted time, and limited visibility into qualified suppliers.
INCONSISTENT SUPPLIER DATA:
Pain Point: 60% report outdated or inaccurate supplier information.
Impact: 93% of supply chain leaders face setbacks due to inaccurate supplier data.
Users cared about capability, credibility, and alignment. They no longer wanted to waste time with suppliers who were unreliable and unfit.
PROJECT GOALS
But what if finding suppliers felt less like a search and more like a query?
No more filling out forms
No more lists to sort through
No more wasted time and effort
I asked myself: What if supplier discovery felt less like a search — and more like a simple ask? What if users could describe what they need, in their own words — and get exactly that? No more forms, no more lists, and no more wasted time.
WHY AI MADE SENSE FOR THIS PROBLEM:
We needed a system that could interpret user intent, surface quality over quantity, and validate supplier details in real time. AI gave us the power to do just that, transforming a manual, uncertain workflow into something more intuitive, efficient, and trustworthy.
SOLUTION
AI - the breakthrough to accelerate supplier discovery
As an avid user of tools like ChatGPT, Claude, Perplexity, and Lovable, I loved the idea of asking a question in plain language and receiving an instant, curated response and not having to dig through endless amounts of information. These tools reduced friction, and it got me wondering:
What if enterprise users could approach supplier discovery in the same way (quick, easy, conversational, and precise)? What if we could eliminate wasted time, outdated search methods, and frustration?
So, I pushed for AI as the solution, steering the conversation away from forms and filters and towards innovation. By combining natural language with AI, we transformed a manual, uncertain process into something more intuitive, efficient, and trustworthy.
COMPETITIVE ANALYSIS
AI Usage
But current platforms are rigid, outdated, and overwhelming
In scoping-out the competition, I found that most competitors were not really providing any relief when it came to supplier discovery. Many of the platforms were directory-style solutions with form-based filtering options (two friction points I already knew existed among users). Instead of presenting qualified, relevant options, these platforms were providing users with a laundry list of suppliers to sort through, leaving users to do all of the work.
I used ChatGPT's 'Deep Research' tool to quickly summarize what competitors like Supplier.io, TealBook, and Thomasnet had to offer, what users loved, and where they struggled. What I found was that most platforms were strict and rigid in how requests could be submitted, offered WAY too many results to sort through, and provided no way to verify supplier data.
SUPPLIER.IO
Large supplier database
Trusted by Fortune 1000 companies
Integrates well with enterprise procurement workflows
Focuses heavily on certifications and not capability
Primarily a search directory
Rigid UI and form-based filters
No real-time guidance
PROS
CONS
TEALBOOK
Real-time supplier data enrichment using AI
Promotes data visibility across procurement systems
Less focus on matchmaking for specific sourcing requests
Can be data-heavy but not sourcing-decision friendly
Less intuitive UX for new sourcing professionals
PROS
CONS
THOMASNET
Huge volume of suppliers, especially in manufacturing
Has category-based filtering
Outdated user experience - clunky and dated UI
No vetting options or compatibility scoring
Lacks user guidance
PROS
CONS
Users were dealing with laundry lists of suppliers to sort through on clunky, rigid, and outdated sites.
To meet our 3-month timeline, I prioritized UI and early testing, helping us ship fast, reduce rework, and get critical feedback while the product was still flexible.
VISUAL DESIGN
Supporting users - from submitting a request to making first contact
As a team, we knew the success of this new product direction would come down to more than just surfacing suppliers — it was about supporting users from start to finish in the sourcing process.
Some of my team members were concerned with how to reduce decision fatigue, guide users through each step, and help them feel confident in the suppliers they ultimately selected. Long lists and clunky workflows had left procurement teams drained and uncertain in the past, and we couldn’t risk repeating those mistakes.
To address these concerns, I focused on designing a guided experience that reduced cognitive load while building confidence at every step.
SUBMITTING A SUPPLIER REQUEST
Users describe what they need in plain language and the system does the heavy lifting. Through AI-powered filtering, only the best-fit suppliers are presented, saving users time and money.
Explore my design process
CURATED SUPPLIER MATCHES
The systems extracts key requirements from the user's prompt and generates tailored matches, helping users focus only on options that meet their specific needs and standards.
Explore my design process
SUPPLIER PROFILES - VERIFYING SUPPLIER DATA
Structured, scannable profiles highlight what matters most. Users have the transparency and confidence they need to make informed decisions.
Explore my design process
This project taught me how to be a Product Thinking Designer, thinking beyond features and focusing on impact.
EDGE CASES
AI Usage
Real-world sourcing isn't straightforward, and neither is user behavior
From vague prompts to not knowing where to start, we knew we had to design for the not-so-great experiences, not just the ideal ones. I leveraged ChatGPT to generate real-world scenarios, identifying areas where users would likely get stuck, confused, or discouraged. I used these insights to build thoughtful safeguards and flexible tools to support users throughout their platform experience.
AVOIDING VAGUE PROMPTS
To guide users in refining their results, we introduced an AI enhancement tool to help refine their prompt without forcing structure.
HELPING USERS FEEL CONFIDENT
I added a ‘Help & Guidance’ section beneath the prompt to make things feel approachable from the start. It gives users prompt examples, explains how the AI-enhancement tool works, and sets clear expectations for what happens after they submit their request, so they feel supported, not unsure.
MY IMPACT
The final product guided, clarified, and delivered confidence
From day one, I was focused on impact. What problem were we solving? What would success look like ? As a product thinker, I helped shape the product's direction. I sat in on early conversations, helped reframe vague ideas into actionable hypotheses, and pushed for clarity around both user needs and business goals.
FRAMED THE PROBLEM, NOT JUST THE SOLUTION
I worked with leadership to clarify who we were designing for, what they actually needed, and how we'd define success.
ADVOCATED FOR SIMPLICITY AND INTENT
I reduced complexity in the supplier request workflow to lower cognitive load and increase usability.
USED UX AS A TOOL FOR BUSINESS ALIGNMENT
My designs weren’t just usable, they reflected key constraints such as limited resourcing, tight timelines, and scalability needs.
DESIGNED FOR OUTCOMES, NOT OUTPUT
Every design decision was tied back to user pain points and product goals, not just visual polish.
FOCUSED ON FLOWS, NOT SCREENS
I mapped full user journeys (including edge cases) to ensure the product worked as a cohesive experience, not a collection of isolated features.
USER TESTING
AI Usage
Integrating User Feedback
We listened closely to what mattered most - what information users needed, what could help them feel confident in making a decision, and what could potentially get in their way. I used ChatGPT to analyze transcripts from live testing sessions and used the results to incorporate the following improvements.
PRODUCT UPDATE #1
BRIDGING MENTAL GAPS
Improving visibility and clarity between extracted requirements and match results.
View Details
PRODUCT UPDATE #2
TAILORING TO OUR USERS
Personalizing the match experience by providing users with additional vetting options.
View Details
PRODUCT UPDATE #3
SHOWING PRODUCT VALUE
Improving UX by giving users a sneak-peek of the benefits of upgrading their subscription.
View Details
NEXT STEPS
Currently in Beta
Sourcing Ally is now in Beta, gathering real-world feedback from enterprise users, validating usability, refining key features, and continuing to align the experience with what procurement teams actually need.
WHAT'S BEING TESTED:
Are users finding matches they can trust?
Is the AI-enhancement tool improving prompt quality?
Are users engaging with supplier profiles and the added vetting option?
REFLECTION
What I'd do differently
The product pivoted twice, and at times I felt a disconnect between what I was designing and why I was designing it. It became clear how easy it is to fall into execution mode when the problem space isn’t fully defined. Next time, I want to lean into product thinking even earlier to help clarify the “why,” define the solution space (as a team), and make sure every design decision is rooted in purpose, not just progress.
Next time, I'll apply product thinking earlier by…
Slowing down to define the core problem
Facilitating workshops to establish alignment on user needs and business goals
Building hypotheses and testing assumptions instead of baking them into the product
Defining success and using it as our North Star
Creating a simple problem statement so we don't lose sight of our mission
What used to take weeks, now takes hours…
…thanks to AI and a smarter way to ask.
What used to take weeks, now takes hours…
…thanks to AI and a smarter way to ask.
CONTEXT
An AI-Powered Sourcing Assistant
An AI-Powered Sourcing Assistant
Finding suppliers shouldn't feel like detective work. Sourcing Ally helps enterprise teams find qualified suppliers by combining natural language with the accuracy and efficiency of artificial intelligence.
Finding suppliers shouldn't feel like detective work. Sourcing Ally helps enterprise teams find qualified suppliers by combining natural language with the accuracy and efficiency of artificial intelligence.
SUCCESS METRICS
40% less effort, 90% match accuracy
40% less effort, 90% match accuracy
Supplier discovery went from weeks to just hours, allowing enterprises to make sourcing decisions 3x faster.
Supplier discovery went from weeks to just hours, allowing enterprises to make sourcing decisions 3x faster.
Let's dive into how this solution came to life!
Let's dive into how this solution came to life!
Full Case Study | 5-10 min read
BACKGROUND
Product History
Product History
Sourcing Ally is the product of Scaleup, a B2B that originated at AWS (Amazon) and went independent in 2023.
Sourcing Ally is the product of Scaleup, a B2B that originated at AWS (Amazon) and went independent in 2023.
Learn more about the product history, solutions that didn't work, and what we learned from them.
Learn More
Learn more about the product history, solutions that didn't work, and what we learned from them.
Learn More
PROJECT DETAILS
My Role
Founding Product Designer
Project Length
3 months
My Team
1 Designer (Me)
1 Project Manager
1 Developer
CEO, CFO, Head of CX, Head of Technology
My Contributions
End-to-End Product Design
AI-Enhanced Workflows
Branding
Marketing Content
PROJECT DETAILS
My Role
Founding Product Designer
Project Length
3 months
My Team
1 Designer (Me)
1 Project Manager
1 Developer
CEO, CFO, Head of CX, Head of Technology
My Contributions
End-to-End Product Design
AI-Enhanced Workflows
Branding
Marketing Content
PROBLEM
An Overwhelming Search for Suppliers
Enterprises face an overwhelming search for qualified suppliers. Endless databases, outdated information, and unreliable data lead to wasted time, costly setbacks, and missed opportunities.
SOLUTION
An AI-Powered Sourcing Assistant
Sourcing Ally simplifies supplier discovery with a natural language prompt - users describe their sourcing needs in plain language and the system interprets, enhances, and delivers high-quality matches.
Learn about some of the challenges we faced during ideation.
Learn More
An Overwhelming Search for Suppliers
PROBLEM
Enterprises face an overwhelming search for qualified suppliers. Endless databases, outdated information, and unreliable data lead to wasted time, costly setbacks, and missed opportunities.
An AI-Powered Sourcing Assistant
SOLUTION
Sourcing Ally simplifies supplier discovery with a natural language prompt - users describe their sourcing needs in plain language and the system interprets, enhances, and delivers high-quality matches.
Learn about some of the challenges we faced during ideation.
Learn More
My Design Approach
Detect
Industry Research
User Insights & Pain Points
Competitive Analysis
Persona Building
Journey Map
Distill
User Flows
Gather UI Inspiration
Define Brand Identity
Design System
UI Design
USER INTERVIEWS
AI Usage
Procurement leaders want quality, not quantity
What we learned was that procurement leaders didn't care about quantity but instead wanted quality. They no longer wanted to waste time with suppliers who were unreliable and unfit.
I used ChatGPT to analyze my interview transcripts, cluster common themes, identify pain points, and summarize feedback. As a result, 3 core pain points emerged:
SUPPLIER SOURCING TAKES TOO LONG:
Pain Point: On average, it takes 57 days just to go from posting a request to finding a supplier.
Impact: Delays operations, increases costs, and limits business agility.
OUTDATED SEARCH METHODS:
Pain Point: 63% search online and 59% rely on magazines, reports, conferences, and word-of-mouth.
Impact: Inconsistent results, wasted time, and limited visibility into qualified suppliers.
INCONSISTENT SUPPLIER DATA:
Pain Point: 60% report outdated or inaccurate supplier information.
Impact: 93% of supply chain leaders face setbacks due to inaccurate supplier data.
Users cared about capability, credibility, and alignment. They no longer wanted to waste time with suppliers who were unreliable and unfit.
My Design Approach
Detect
Industry Research
User Insights & Pain Points
Competitive Analysis
Persona Building
Journey Map
Distill
User Flows
Gather UI Inspiration
Define Brand Identity
Design System
UI Design
Design
Prototypes
User Testing
Implement Feedback
Launch
USER INTERVIEWS
Synthesized Using AI
Procurement leaders want quality, not quantity
What we learned was that procurement leaders didn't care about quantity but instead wanted quality. They no longer wanted to waste time with suppliers who were unreliable and unfit.
I used ChatGPT to analyze my interview transcripts, cluster common themes, identify pain points, and summarize feedback. As a result, 3 core pain points emerged:
PAIN POINT #1 - Supplier Sourcing takes WAY too long
Pain Point: On average, it takes 57 days just to go from posting a request to finding a supplier.
Impact: Delays operations, increases costs, and limits business agility.
PAIN POINT #2 - Outdated Search Methods
Pain Point: 63% search online and 59% rely on magazines, reports, conferences, and word-of-mouth.
Impact: Inconsistent results, wasted time, and limited visibility into qualified suppliers.
PAIN POINT #3 - Inconsistent Supplier Data
Pain Point: 60% report outdated or inaccurate supplier information.
Impact: 93% of supply chain leaders face setbacks due to inaccurate supplier data.
Users cared about capability, credibility, and alignment. They no longer wanted to waste time with suppliers who were unreliable and unfit.
SUPPLIER.IO
Large supplier database
Trusted by Fortune 1000 companies
Integrates well with enterprise procurement workflows
Focuses heavily on certifications and not capability
Primarily a search directory
Rigid UI and form-based filters
No real-time guidance
PROS
CONS
TEALBOOK
Real-time supplier data enrichment using AI
Promotes data visibility across procurement systems
Less focus on matchmaking for specific sourcing requests
Can be data-heavy but not sourcing-decision friendly
Less intuitive UX for new sourcing professionals
PROS
CONS
THOMASNET
Huge volume of suppliers, especially in manufacturing
Has category-based filtering
Outdated user experience - clunky and dated UI
No vetting options or compatibility scoring
Lacks user guidance
PROS
CONS
COMPETITIVE ANALYSIS
Summarized Using AI
But current platforms are rigid, outdated, and overwhelming
But current platforms are rigid, outdated, and overwhelming
In scoping-out the competition, I found that most competitors were not really providing any relief when it came to supplier discovery. Many of the platforms were directory-style solutions with form-based filtering options (two friction points I already knew existed among users). Instead of presenting qualified, relevant options, these platforms were providing users with a laundry list of suppliers to sort through, leaving users to do all of the work.
I used ChatGPT's 'Deep Research' tool to quickly summarize what competitors like Supplier.io, TealBook, and Thomasnet had to offer, what users loved, and where they struggled. What I found was that most platforms were strict and rigid in how requests could be submitted, offered WAY too many results to sort through, and provided no way to verify supplier data.
In scoping-out the competition, I found that most competitors were not really providing any relief when it came to supplier discovery. Many of the platforms were directory-style solutions with form-based filtering options (two friction points I already knew existed among users). Instead of presenting qualified, relevant options, these platforms were providing users with a laundry list of suppliers to sort through, leaving users to do all of the work.
I used ChatGPT's 'Deep Research' tool to quickly summarize what competitors like Supplier.io, TealBook, and Thomasnet had to offer, what users loved, and where they struggled. What I found was that most platforms were strict and rigid in how requests could be submitted, offered WAY too many results to sort through, and provided no way to verify supplier data.
SUPPLIER.IO
Large supplier database
Trusted by Fortune 1000 companies
Integrates well with enterprise procurement workflows
Focuses heavily on certifications and not capability
Primarily a search directory
Rigid UI and form-based filters
No real-time guidance
PROS
CONS
TEALBOOK
Real-time supplier data enrichment using AI
Promotes data visibility across procurement systems
Less focus on matchmaking for specific sourcing requests
Can be data-heavy but not sourcing-decision friendly
Less intuitive UX for new sourcing professionals
PROS
CONS
THOMASNET
Huge volume of suppliers, especially in manufacturing
Has category-based filtering
Outdated user experience - clunky and dated UI
No vetting options or compatibility scoring
Lacks user guidance
PROS
CONS
Users were dealing with laundry lists of suppliers to sort through on clunky, rigid, and outdated sites.
Therefore, enterprises face delays, high costs, and missed opportunities
As a result, enterprise users were stuck in the past - relying on slow, labor-intensive processes that were not only draining their resources and their time, but were preventing them from finding the right suppliers at the right time.
PROJECT GOALS
But what if finding suppliers felt less like a search and more like a query?
I asked myself: What if supplier discovery felt less like a search — and more like a simple ask? What if users could describe what they need, in their own words — and get exactly that? No more forms, no more lists, and no more wasted time.
No more filling out forms
No more laundry lists to sort through
No more wasted time and effort
SOLUTION
AI - the breakthrough to accelerate supplier discovery
As an avid user of tools like ChatGPT, Claude, Perplexity, and Lovable, I loved the idea of asking a question in plain language and receiving an instant, curated response and not having to dig through endless amounts of information. These tools reduced friction, and it got me wondering:
What if enterprise users could approach supplier discovery in the same way (quick, easy, conversational, and precise)? What if we could eliminate wasted time, outdated search methods, and frustration?
So, I pushed for AI as the solution, steering the conversation away from forms and filters and towards innovation. By combining natural language with AI, we transformed a manual, uncertain process into something more intuitive, efficient, and trustworthy.
WHY AI MADE SENSE FOR THIS PROBLEM:
We needed a system that could interpret user intent, surface quality over quantity, and validate supplier details in real time. AI gave us the power to do just that, transforming a manual, uncertain workflow into something more intuitive, efficient, and trustworthy.
Therefore, enterprises face delays, high costs, and missed opportunities
As a result, enterprise users were stuck in the past - relying on slow, labor-intensive processes that were not only draining their resources and their time, but were preventing them from finding the right suppliers at the right time.
PROJECT GOALS
But what if finding suppliers felt less like a search, and more like a query?
I asked myself: What if supplier discovery felt less like a search — and more like a simple ask? What if users could describe what they need, in their own words — and get exactly that? No more forms, no more lists, and no more wasted time.
No more filling out forms
No more laundry lists to sort through
No more wasted time and effort
SOLUTION
AI - the breakthrough to accelerate supplier discovery
As an avid user of tools like ChatGPT, Claude, Perplexity, and Lovable, I loved the idea of asking a question in plain language and receiving an instant, curated response and not having to dig through endless amounts of information. These tools reduced friction, and it got me wondering:
What if enterprise users could approach supplier discovery in the same way (quick, easy, conversational, and precise)? What if we could eliminate wasted time, outdated search methods, and frustration?
So, I pushed for AI as the solution, steering the conversation away from forms and filters and towards innovation. By combining natural language with AI, we transformed a manual, uncertain process into something more intuitive, efficient, and trustworthy.
WHY AI MADE SENSE FOR THIS PROBLEM:
We needed a system that could interpret user intent, surface quality over quantity, and validate supplier details in real time. AI gave us the power to do just that, transforming a manual, uncertain workflow into something more intuitive, efficient, and trustworthy.
To meet our 3-month timeline, I prioritized UI and early testing, helping us ship fast, reduce rework, and get critical feedback while the product was still flexible.
To meet our 3-month timeline, I prioritized UI and early testing, helping us ship fast, reduce rework, and get critical feedback while the product was still flexible.
VISUAL DESIGN
Supporting users - from submitting a request to making first contact
As a team, we knew the success of this new product direction would come down to more than just surfacing suppliers — it was about supporting users from start to finish in the sourcing process.
Some of my team members were concerned with how to reduce decision fatigue, guide users through each step, and help them feel confident in the suppliers they ultimately selected. Long lists and clunky workflows had left procurement teams drained and uncertain in the past, and we couldn’t risk repeating those mistakes.
To address these concerns, I focused on designing a guided experience that reduced cognitive load while building confidence at every step.
SUBMITTING A SUPPLIER REQUEST
Users describe what they need in plain language and the system does the heavy lifting. Through AI-powered filtering, only the best-fit suppliers are presented, saving users time and money.
CURATED SUPPLIER MATCHES
The systems extracts key requirements from the user's prompt and generates tailored matches, helping users focus only on options that meet their specific needs and standards.
SUPPLIER PROFILES - VERIFYING SUPPLIER DATA
Structured, scannable profiles highlight what matters most. Users have the transparency and confidence they need to make informed decisions.
VISUAL DESIGN
Supporting users - from submitting a request to making first contact
As a team, we knew the success of this new product direction would come down to more than just surfacing suppliers — it was about supporting users from start to finish in the sourcing process.
Some of my team members were concerned with how to reduce decision fatigue, guide users through each step, and help them feel confident in the suppliers they ultimately selected. Long lists and clunky workflows had left procurement teams drained and uncertain in the past, and we couldn’t risk repeating those mistakes.
To address these concerns, I focused on designing a guided experience that reduced cognitive load while building confidence at every step.
SUBMITTING A SUPPLIER REQUEST
Users describe what they need in plain language and the system does the heavy lifting. Through AI-powered filtering, only the best-fit suppliers are presented, saving users time and money.
CURATED SUPPLIER MATCHES
The systems extracts key requirements from the user's prompt and generates tailored matches, helping users focus only on options that meet their specific needs and standards.
SUPPLIER PROFILES - VERIFYING SUPPLIER DATA
Structured, scannable profiles highlight what matters most. Users have the transparency and confidence they need to make informed decisions.
USER TESTING
Summarized Using AI
USER TESTING
Synthesized Using AI
Integrating User Feedback
Integrating User Feedback
We listened closely to what mattered most - what information users needed, what could help them feel confident in making a decision, and what could potentially get in their way. I used ChatGPT to analyze transcripts from live testing sessions and used the results to incorporate the following improvements.
We listened closely to what mattered most - what information users needed, what could help them feel confident in making a decision, and what could potentially get in their way. I used ChatGPT to analyze transcripts from live testing sessions and used the results to incorporate the following improvements.
PRODUCT UPDATE #1
BRIDGING MENTAL GAPS
Improving visibility and clarity between extracted requirements and match results.
View Details
PRODUCT UPDATE #2
TAILORING TO OUR USERS
Personalizing the match experience by providing users with additional vetting options.
View Details
PRODUCT UPDATE #3
SHOWING PRODUCT VALUE
Improving UX by giving users a sneak-peek of the benefits of upgrading their subscription.
View Details
PRODUCT UPDATE #1
BRIDGING MENTAL GAPS
Improving visibility and clarity between extracted requirements and match results.
View Details
PRODUCT UPDATE #2
TAILORING TO OUR USERS
Personalizing the match experience by providing users with additional vetting options.
View Details
PRODUCT UPDATE #3
SHOWING PRODUCT VALUE
Improving UX by giving users a sneak-peek of the benefits of upgrading their subscription.
View Details
EDGE CASES
Enhanced Using AI
EDGE CASES
Enhanced Using AI
Real-world sourcing isn't straightforward, and neither is user behavior
Real-world sourcing isn't straightforward, and neither is user behavior
From vague prompts to not knowing where to start, we knew we had to design for the not-so-great experiences, not just the ideal ones. I leveraged ChatGPT to generate real-world scenarios, identifying areas where users would likely get stuck, confused, or discouraged. I used these insights to build thoughtful safeguards and flexible tools to support users throughout their platform experience.
From vague prompts to not knowing where to start, we knew we had to design for the not-so-great experiences, not just the ideal ones. I leveraged ChatGPT to generate real-world scenarios, identifying areas where users would likely get stuck, confused, or discouraged. I used these insights to build thoughtful safeguards and flexible tools to support users throughout their platform experience.
AVOIDING VAGUE PROMPTS
AVOIDING VAGUE PROMPTS
To guide users in refining their results, we introduced an AI enhancement tool to help refine their prompt without forcing structure.
To guide users in refining their results, we introduced an AI enhancement tool to help refine their prompt without forcing structure.
HELPING USERS FEEL CONFIDENT FROM THE START
HELPING USERS FEEL CONFIDENT FROM THE START
I added a ‘Help & Guidance’ section beneath the prompt to make things feel approachable from the start. It gives users prompt examples, explains how the AI-enhancement tool works, and sets clear expectations for what happens after they submit their request, so they feel supported, not unsure.
I added a ‘Help & Guidance’ section beneath the prompt to make things feel approachable from the start. It gives users prompt examples, explains how the AI-enhancement tool works, and sets clear expectations for what happens after they submit their request, so they feel supported, not unsure.
MY IMPACT
The final product guided, clarified, and delivered confidence
From day one, I was focused on impact. What problem were we solving? What would success look like ? As a product thinker, I helped shape the product's direction. I sat in on early conversations, helped reframe vague ideas into actionable hypotheses, and pushed for clarity around both user needs and business goals.
FRAMED THE PROBLEM, NOT JUST THE SOLUTION
I worked with leadership to clarify who we were designing for, what they actually needed, and how we'd define success.
ADVOCATED FOR SIMPLICITY AND INTENT
I reduced complexity in the supplier request workflow to lower cognitive load and increase usability.
USED UX AS A TOOL FOR BUSINESS ALIGNMENT
My designs weren’t just usable, they reflected key constraints such as limited resourcing, tight timelines, and scalability needs.
DESIGNED FOR OUTCOMES, NOT OUTPUT
Every design decision was tied back to user pain points and product goals, not just visual polish.
FOCUSED ON FLOWS, NOT SCREENS
I mapped full user journeys (including edge cases) to ensure the product worked as a cohesive experience, not a collection of isolated features.
This project taught me how to be a Product Thinking Designer, thinking beyond features and focusing on impact.
This project taught me how to be a Product Thinking Designer, thinking beyond features and focusing on impact.
MY IMPACT
The final product guided, clarified, and provided confidence
From day one, I was focused on impact. What problem were we solving? What would success look like ? As a product thinker, I helped shape the product's direction. I sat in on early conversations, helped reframe vague ideas into actionable hypotheses, and pushed for clarity around both user needs and business goals.
FRAMED THE PROBLEM, NOT JUST THE SOLUTION
I worked with leadership to clarify who we were designing for, what they actually needed, and how we'd define success.
ADVOCATED FOR SIMPLICITY AND INTENT
I reduced complexity in the supplier request workflow to lower cognitive load and increase usability.
USED UX AS A TOOL FOR BUSINESS ALIGNMENT
My designs weren’t just usable, they reflected key constraints such as limited resourcing, tight timelines, and scalability needs.
DESIGNED FOR OUTCOMES, NOT OUTPUT
Every design decision was tied back to user pain points and product goals, not just visual polish.
FOCUSED ON FLOWS, NOT SCREENS
I mapped full user journeys (including edge cases) to ensure the product worked as a cohesive experience, not a collection of isolated features.
REFLECTION
What I'd do differently
What I'd do differently
The product pivoted twice, and at times I felt a disconnect between what I was designing and why I was designing it. It became clear how easy it is to fall into execution mode when the problem space isn’t fully defined. Next time, I want to lean into product thinking even earlier to help clarify the “why,” define the solution space (as a team), and make sure every design decision is rooted in purpose, not just progress.
The product pivoted twice, and at times I felt a disconnect between what I was designing and why I was designing it. It became clear how easy it is to fall into execution mode when the problem space isn’t fully defined. Next time, I want to lean into product thinking even earlier to help clarify the “why,” define the solution space (as a team), and make sure every design decision is rooted in purpose, not just progress.
Next time, I'll apply product thinking earlier by…
Next time, I'll apply product thinking earlier by…
Slowing down to define the core problem
Facilitating workshops to establish alignment on user needs and business goals
Building hypotheses and testing assumptions instead of baking them into the product
Defining success and using it as our North Star
Creating a simple problem statement so we don't lose sight of our mission
Slowing down to define the core problem
Facilitating workshops to establish alignment on user needs and business goals
Building hypotheses and testing assumptions instead of baking them into the product
Defining success and using it as our North Star
Creating a simple problem statement so we don't lose sight of our mission
NEXT STEPS
Currently in Beta
Currently in Beta
Sourcing Ally is now in Beta, gathering real-world feedback from enterprise users, validating usability, refining key features, and continuing to align the experience with what procurement teams actually need.
Sourcing Ally is now in Beta, gathering real-world feedback from enterprise users, validating usability, refining key features, and continuing to align the experience with what procurement teams actually need.
WHAT'S BEING TESTED:
WHAT'S BEING TESTED:
Are users finding matches they can trust?
Is the AI-enhancement tool improving prompt quality?
Are users engaging with supplier profiles and the added vetting option?
Are users finding matches they can trust?
Is the AI-enhancement tool improving prompt quality?
Are users engaging with supplier profiles and the added vetting option?