01.Workp. 1 02.Aboutp. 2 03.Resumep. 3 pick a page, any page →
ask "why" until it gets uncomfortable. then once more.
field notes47 interviews
9 usability studies
1 framework
field notes of a researcher who started as a storyteller
UX Researcher · Seattle

I find the human signal
in complex systems

I'm Karthik Srinivasan. I started as a software engineer. Research taught me to listen. Now I help enterprise teams build things people actually understand, and want to use.

keep reading
← start here
good research ships. great research changes the roadmap.
Selected work

Research that shipped

Two case studies from Equinix. Research shaped these products from the first question to launch day, and kept shaping them after.

Case 01
Solution Builder network topology diagram
Enterprise SaaS · Equinix
Solution Builder

Designing collaborative network infrastructure solutioning. From a broken pre-sales process to a tool that sells itself.

3,000
MAU
2 days
saved / project
14%
more revenue
InterviewsKano SurveyConcept TestingNPSAmplitude
Read case study →
Case 02
Equinix Customer Portal with unified navigation
Information Architecture · Equinix
Equinix Portal

Unifying 12 disconnected product portals into one coherent experience, built on the mental models of the customers using them.

71%
revenue inc.
18%
lower churn
80
card sort ptcpts
Card SortingUsability TestingInterviewsNPS
Read case study →
What I believe

Research isn't a step.
It's the through-line
from question to impact.

"designing without research
results in a jarred experience"
Start with listening

I go where users work, not where they present. The gap between what people say and what they do is where real insight lives.

1
Pick the right method

A Kano survey with 31 stakeholders or a 75-minute interview with one expert. The best research answers the question efficiently.

2
Make findings move people

Insights that sit in a report change nothing. I build narratives that make teams want to act.

3
begin with chapter one: Solution Builder →
new domain,
new questions

Solution Builder

Case Study · Equinix · 2022–2023

A solutioning process held together by spreadsheets becomes the canvas 3,000 people use every day.

Before this tool existed, designing a network solution meant days of back-and-forth, fragile spreadsheets, and no shared canvas. I was asked to find out why, then help build something better.

My role
Lead UX Researcher
Duration
~12 months
Methods
Interviews · Kano · Concept Testing · NPS
Solution Builder interface
Solution Builder · canvas view
4-day sprint tight deadline
The problem

A pre-sales process nobody could defend

GSAs spent 2+ days on manual solution design before a single customer conversation could happen. I volunteered for a 4-day sprint. Tight deadline, real stakes, a chance to prove research could move fast and still matter.

"I spend two days just getting to the point where I can show a customer what a solution looks like. And then they change something and I start over."

GSA interview participant

Networking was entirely new to me. I had to learn new concepts and ask good questions without pretending expertise. My CS background helped me go deep technically; my HCI training kept me curious about the human layer.

🕐
2 days lost before any conversation
Manual design prep consumed time that should have been spent with customers.
🔀
No shared canvas
Customers and architects had no common visualization. Every change meant email chains.
📉
Missed upsell moments
Without visibility into existing deployments, natural upsell opportunities were invisible.
interviews → survey → clarity
Phase One

Understanding the problem space

Interviews, a stakeholder workshop, and a Kano survey. Three methods, one purpose: replace assumption with evidence.

1
Customer & Stakeholder Interviews
Mapped GSA and customer workflows in parallel. Short deadline forced sharp, purposeful questions and fast synthesis.
2
JTBD Workshop (n=9 stakeholders)
Translated raw insights into a feature list collaboratively. Got competing stakeholders looking at the same evidence.
3
Kano Survey (n=31 GSAs)
Distinguished must-haves from nice-to-haves with statistical clarity. Ended the feature debate overnight.
what the Kano survey revealed
Must-haves
Pricing & quoting · Deployment visibility · Latency metrics
Nice-to-have
Collaboration · Onboarding tutorials · Templates
ended the feature debate
polarized reactions = the insight
Phase Two

Testing the solution

Two rounds of concept testing. What the second round revealed wasn't a problem. It was the insight that changed the design direction entirely.

🏗️
Round 1 — 6 GSAs
Detailed technical feedback made navigation and diagramming significantly more intuitive before customers ever saw it.
⚡
Round 2 — 7 customers: polarized
Experienced engineers loved the freedom. Novice users were lost. This wasn't failure. It was the insight that led to two distinct user flows.

"This is exactly what I needed. I can finally see the whole solution in one place."

Experienced GSA — Round 1

"I'm not sure where to start. There's so much here."

Novice user — Round 2 · the insight that changed the design

Beta 4 months after concept testing. GA 6 months later.

two flows: guided + expert mode
NPS 43% vs 21% avg
Phase Three

Post-launch measurement

Launch was the beginning, not the end. Attitudinal (NPS) and behavioral (Amplitude) signals combined.

NPS Survey (n=631)

Pearson correlation identified which satisfaction drivers most strongly predicted recommendation likelihood.

43%
NPS Score
vs. 21% industry avg
Click interaction tracking
Every node addition, topology change tracked via Amplitude to identify friction continuously.
Customer journey funnels
Analyzed paths from topology view to quote generation, pinpointing exactly where users dropped off.
Feature relevance scoring
Action frequency distinguished real engagement from features quietly ignored post-launch.
the research made this happen
Outcomes

What the research made possible

—
Monthly Active UsersCustomers and Solution Architects using the tool daily
2 days
Saved per projectGSAs reported two full days back on every pre-sales cycle
—
More customer spendingEstimated revenue uplift from improved solutioning

"The success of Solution Builder led to our VP enlisting me in conversations for a massive product overhaul, from the very first meeting."

Karthik's reflection
Shipped ✓
next case: Equinix Portal →
12 portals, 12 mental models

Equinix Portal

Case Study · Equinix · 2023–2024

Twelve portals, twelve mental models, and the research that earned one navigation customers finally trusted.

12 portals for 12 products. Customers navigated by guessing. My job: find out how customers actually think about these products, then make the new IA match their mental model, not the org chart.

My role
Lead UX Researcher
Scope
12 portals unified
Methods
Card Sorting · Usability Testing
Equinix Portal with unified navigation
Equinix Portal · unified nav
customer voice must be undeniable
The challenge

12 portals. 12 mental models. Zero coherence.

Every product had its own portal, its own navigation, its own logic. And many internal stakeholders feared a unified portal would reduce their product's visibility. Research had to be the neutral arbiter.

🗂️
12 separate portals
Billing, orders, and support lived in different places for every product.
🤔
Jargon as navigation
"IBX" and "Packet" were internal names that meant nothing to paying customers.
⚡
High political stakes
The customer voice needed to be undeniable, not just compelling.
1,000+ customers contacted manually
Phase One

Mapping how customers actually think

The largest hybrid card sort in the company's research history. 80 participants, two rounds, one clear picture of how customers group the products they use.

study design
Round 1 — Internal Usersn=50
72 cards · 7 predefined categories · Hybrid card sort over Optimal Sort
Round 2 — Customersn=30
25 cards · Focused on ambiguous groupings · Similarity scores + dendrogram
manually compiled 1,000+ customer records for recruitment

What the cards revealed

1
Rename "IBX" → "Colocation"
Many couldn't place "IBX." "Colocation" was immediately understood by nearly everyone.
2
Rename "Packet" → "Equinix Metal"
Packet was frequently sorted into "I don't know." Customers didn't connect it to Equinix at all.
3
Functional names win
"Manage Cables," "Install Equipment" had far higher agreement rates than branded names.
NPS 81% vs 43% industry
Phase Two

Testing the new navigation model

Prototype usability testing with 12 customers before anything shipped. The results were unambiguous.

✅
"Organized" and "modern"
Participants compared the new left nav favorably to AWS and Azure.
🔍
Discovery happened naturally
Users found products like Virtual Services for the first time. The nav became a product catalog.
🎯
Consistent naming confirmed
"Colocation" used consistently across nav, marketing, and content created a coherent brand identity.
Post-launch NPS (n=251)
81%
vs. 43% industry standard

Pearson r=0.68. Ease of navigation strongly correlated with recommendation likelihood.

collaborated with data analyst
unified portal customers trusted
Business impact

A unified portal customers trusted

—
Revenue increaseNetwork Edge gained revenue after the unified portal launched
—
Lower churnImproved navigation reduced cumulative churn across all products
3
Products rebrandedColocation, Internet Access, Equinix Metal. All renamed from what customers showed us.
Field-tested ✓
how I learned to work this way →
every product is a story someone is stuck inside
✨ happier endings →
research log3 publications
1 framework (AURA)
still listening
storyteller
first software
dev → HCI 6+ years
enterprise

I started as an engineer.
Research taught me to listen.

Before I was a researcher, I was a storyteller. As a child, family and friends would ask me, "Tell us some story", and I would narrate tales, reading the eyes and expressions of the room, knowing when to dramatize, when to tone down. I paid close attention to what moved people.

Later, as a screenwriter, I used ethnography and observation to build characters. I wanted them to be real. Stories blossomed out of what I saw in the world. One day, a writer friend told me my stories needed happier endings. That question stayed with me: how do you actually rewrite someone's story to give them a better outcome?

It struck me that being a UX Researcher does exactly that. I help rewrite the small inconveniences and major frustrations people face with technology, and give them happier endings. That realization pulled me fully into research. I brought the same curiosity I had as a storyteller, the same close attention to people, and the same drive to understand what was really going on beneath the surface.

Now I work at the intersection of complex enterprise systems and the humans who use them. Cloud infrastructure, data centers, edge computing. The products are technical. The people using them are deeply human. That gap is where I work.

If I had to distill six years of research into one sentence it would be this: the best insight is the one that makes the room go quiet, because everyone in it suddenly sees what they couldn't see before.

Connect on LinkedIn
the plot so far ↓
act structureact i · engineer
act ii · storyteller
act iii · researcher
"your stories need happier endings"
the note that changed everything
Career journey

A path built on curiosity

↓ Full resume (PDF)
Jun 2025 – Present
UX Researcher & Author
Independent · Published Researcher
where AURA started

Three peer-reviewed papers submitted to HAI 2026 and NordiCHI 2026, on AI-driven classification tools, IAM permission recommendations, and AURA, a new framework for measuring usability in agentic AI systems.

Nov 2020 – Dec 2024
Senior UX Researcher
Equinix
four years, 2 → 8 researchers

Led research across cloud, networking, and AI products. Built the research repository from scratch, grew the team from 2 to 8, and shaped product roadmaps at VP and C-suite level.

Sep 2019 – Sep 2020
UX Researcher (Contract)
Microsoft · Azure Cloud & AI / Edge Browser
the year that taught me scale

Behavioral and attitudinal research for Azure and Edge. Edge Browser work contributed to a 46% adoption increase post-launch. Azure work uncovered 30+ new design opportunities.

Jan 2018 – May 2019
UX Research Associate
CAT Lab · Georgia Tech
fell in love with the "why"

Longitudinal research on gamification's impact on engagement in MOOCs, advised by Dr. Betsy DiSalvo. Thesis recognized as runner-up best project in the cohort.

methods are tools. curiosity is the craft.
mixed methods > either/or
Capabilities

How I work

Research methods
  • Interviewing & Contextual Inquiry
  • Usability Testing
  • Concept Testing
  • Card Sorting
  • Kano Survey & NPS
  • A/B Testing & Benchmarking
  • Co-creation Activities
  • Longitudinal Studies
Tools & analysis
  • Figma, Miro, Cursor
  • Amplitude, Google Analytics
  • Power BI, Tableau
  • SPSS, Dovetail
  • UserZoom, UserTesting
  • Qualtrics, Optimal Sort
  • SQL, Python, JavaScript
Leadership
  • Executive storytelling
  • Research repository building
  • Team building & mentoring
  • Research culture & ops
  • Cross-functional facilitation
  • Stakeholder management
What I believe

Designing without research results in a jarred experience for users.

1
Pushing outside comfort rewards immensely

Networking was entirely new when I joined Equinix. I had to learn and ask without pretending expertise. That discomfort is exactly where the best research happens.

2
Complexity is not an excuse for bad experiences

Enterprise software is hard. But "it's complex" is never a reason to leave users confused. Complexity is the brief, not the barrier.

3
Research has to move at the speed of business

I've run complete research cycles in 4 days. Speed doesn't sacrifice rigour. It forces sharper choices about what matters most.

4
The customer voice must be undeniable

When stakeholders resist change, data isn't always enough. The right quote from the right user, heard directly, changes rooms.

5
Research is storytelling, with a happier ending

I started as a screenwriter before becoming a researcher. I wanted my characters to be real. Stories blossomed out of what I saw in the world. A writer friend once said my stories needed happier endings. It struck me that UX Research is exactly that: rewriting the small inconveniences and major frustrations people face with technology, and giving them better outcomes.

the formal version, for the record →
Resume

Karthik Srinivasan

UX Researcher · 6+ years in enterprise technology · Cloud, AI & Security products

Passionate about transforming enterprise technologies into human-centered experiences that influence product strategy and drive real user impact, with deep expertise spanning cloud infrastructure, AI systems, and security products.

↓ Download resume (PDF)
Experience
UX Researcher & Author
Independent
Jun 2025 – Present
  • Uncovered structural gaps in how developers interact with AI-generated cloud security recommendations and how enterprise AI classification tools fail trained users, producing design principles and frameworks across three peer-reviewed papers submitted to HAI 2026 and NordiCHI 2026.
  • Proposed AURA, a new framework for measuring usability in agentic AI systems, addressing a measurement gap no existing instrument had covered for AI agents, chatbots, and generative AI systems.
Senior UX Researcher
Equinix
Nov 2020 – Dec 2024
  • Owned end-to-end research strategy across enterprise cloud, networking & AI products, partnering with product, design, marketing & engineering teams to shape roadmaps and drive decisions at the product and executive level.
  • Drove multi-method generative research that uncovered key market opportunities, leading to a new product launch and 10% growth in customers.
  • Elevated customer satisfaction by 18% across Edge computing products by designing evaluative studies, post-launch analytics and benchmarks that identified usability and task efficiency gaps before they reached scale.
  • Mentored junior UX researchers, standardized onboarding processes, and built a robust research repository from the ground up across UX teams.
User Researcher (Contract)
Microsoft
Sep 2019 – Sep 2020
  • Azure Cloud & AI Team | Apr – Sep 2020
  • Led mixed-method research studies for Azure Cloud & AI, evaluating new designs, patterns and user workflows to improve product usability.
  • Influenced product decisions through quantitative studies that uncovered 30+ new design opportunities, leading to an enhanced portal experience.
  • Edge Browser Team | Sep 2019 – Feb 2020
  • Conducted user behavior and attitudinal studies to optimize first-run experiences, achieving 46% increase in Edge browser adoption post-launch.
  • Translated research findings into strategic narratives that drove alignment among product managers, designers & engineers around key design decisions.
UX Research Associate
CAT Lab · Georgia Tech
Jan 2018 – May 2019
  • Conducted longitudinal research examining how gamification in education influenced young adult engagement and course completion in MOOCs, informing early product recommendations.
Education
MS, Human-Computer Interaction
Georgia Institute of Technology
GPA: 4.0 / 4.0
BTech, Computer Science
SRM University, India
GPA: 9.28 / 10
Skills & Tools
InterviewingContextual InquiryFocus GroupCo-creationUsability TestingConcept TestingA/B TestingBenchmarkingSurveyCard SortingAmplitudeGoogle AnalyticsPower BISPSSFigmaMiroSQLPythonJavaScript
Publications
Beyond SUS: A Systematic Review of Usability Measurement in AI Systems and the AURA Framework
HAI 2026 (submitted)
Understanding AI-Generated IAM Permission Recommendations: Learnability, Reliance, and Consequence Visibility among Cloud Developers
HAI 2026 (submitted)
Designing for Learnability and Trust in AI Data Classification
NordiCHI 2026 (submitted)
the full version

Download the complete PDF version.

↓ Download PDF