Fewer repeat contacts at Conda
Turned support data into product fixes, so the same problems stopped coming back to the inbox.
Support operations + the build
I fix how customer support runs, then set up or build the tools it runs on.
For the team with macros nobody owns, answers in one person’s head, or a spreadsheet doing a job it has outgrown.
Tell me where it gets stuckZendesk · Intercom · Knowledge bases · Internal tools
The work usually starts here
“Which question does your team answer every single day?”
Read the tickets. Follow the handoffs. Look at what the team measures.
Clear routing, useful macros and answers people can find.
Configure the tool, or build a small one. Leave a written handoff.
Turned support data into product fixes, so the same problems stopped coming back to the inbox.
Rebuilt the Help Center and launched the first public status page.
How I can help
Zendesk or Intercom reviewed and cleaned up. Leave with a ranked fix list, or have the fixes applied.
SOPs, a knowledge base and new-agent onboarding built from conversations with the people doing the work.
A scoped pilot for policy and how-to questions, with rules for what the AI can promise and when a person takes over.
Someone to own the helpdesk, weekly reporting, QA and escalations as the team grows.
An admin view, intake workflow or small internal tool on Next.js and Supabase, scoped around an operational problem. The runbook comes with it.
Built as well as planned
Bookings, payments, messaging and an admin back office, built with AI coding tools.
The operations were part of the work too: guide onboarding, payouts, refunds and disputes, written up in a runbook.
A place to start
From the first message to closed. Include the tool you use and the part that gets stuck, and I’ll tell you where I’d start.
Email about your support