I turn customer problems into business opportunities.

Customers were skimming through long bot messages and missing the next step. I simplified the copy and flow so each message asked for one clear action. Completion rose from 85% to 91%, and curators could see who needed a follow-up.

The problem

New customers struggled to complete onboarding; bot messages were too easy to skim past.

What I did

Reviewed the friction points, rewrote the bot copy and step logic, and helped surface stalled onboarding to curators.

What changed

Completion increased from 85% to 91%.

What it taught me

Clear language is part of product design, not a final polish step.

Customer information lived in different systems, which made even simple questions slow to answer. I built a shared customer platform in Lovable, linked Amplitude and internal systems through APIs, and made key customer signals easier to find. Up to ten colleagues now use the shared card.

The problem

Fragmented data slowed down the support and expert teams.

My role

I raised the issue; after early spreadsheet attempts, my manager and I agreed to build an internal platform. I built the dashboard, customer card and coach report in Lovable, connected data through APIs, configured Salebot tags and updated AI-generated customer summaries.

What changed

A shared workspace for up to ten colleagues and an estimated ≈30% reduction in lookup time. The team could see onboarding drop-off by plan and potential churn risks for the first time.

Honest caveat

The 30% is my estimate, not a measured time study.

Payment reports didn’t add up. I traced the discrepancies across systems and found that auto-renewal settings were calculating subscription end dates incorrectly. That discovery exposed billing delays and extra credited days — and helped prevent ≈$25K a year in revenue leakage.

The problem

Repeated discrepancies between GetCourse and internal reports.

My role

Investigated how the systems interact, isolated the issues, explained the technical causes and handed the fixes to the responsible team.

What changed

≈$25K a year in revenue leakage prevented.

Customer complaints revealed a gap between what people thought they were buying and what the subscription actually included. My manager and I reviewed two years of new-customer retention data. I proposed separating Program and Subscription, developed pricing logic, and worked with marketing and sales to align the message and train the team.

The problem

Expectation gaps created avoidable confusion.

My role

Identified the recurring complaint theme, analyzed the data with my manager, proposed the offer structure and pricing logic, helped align with the head of marketing and sales, and trained the sales team.

What changed

Revised offers and sales messaging.

Some customers need support before they think to ask for it. I launched a bot-based Health Check in one program to learn who was struggling. About half responded; the check-in showed that 20% were falling behind. Coaches and support used the responses to guide follow-up.

How do you help someone choose their next step without overwhelming them? I proposed two concepts, Diagnostic and Navigator, and built clickable prototypes in Claude. The CEO and CPO own the underlying methodology; my role is to translate it into a usable product experience.