We coached the founder of a leadership-coaching practice to make AI a dependable part of how she runs her business. She now drafts proposals, client contracts, and presentations through Claude skills she owns, instead of rebuilding inconsistent AI output by hand.
At a glance
The challenge
Honeycomb Teams is a leadership and team-coaching practice, and its founder already used Claude every day for proposals, contracts, and presentations. The output was inconsistent, though, and she often rebuilt it by hand, especially the on-brand slide decks, while the rest of her AI work sat in chat windows she could not reliably reproduce. She wanted to understand what was happening under the hood and build the competence herself, so she could advance the capabilities of her business further. She saw an opportunity to gain much of her time back from operational aspects of her business by utilizing AI. That way, she can focus on creating more value for her clients.
What we built together
- Crafted a coaching infrastructure so that preparation and monitoring progress on engagements is a streamlined, consistent process
- Moved her AI work out of one-off chats into Claude skills and Projects she owns, packaging her proprietary frameworks for reuse
- Consolidated her scattered Claude projects into one organized workspace for running her business
- Built a voice from her own writing, with brand controls so every output sounds like Honeycomb
- Designed guardrails for her client data and IP, with model training turned off
- Taught the fundamentals: matching the model, mode, and effort to the task
The results
- She drafts proposals, contracts, and presentations through AI she owns, with consistency and ease
- Engineering templates for contracts, proposals, and team effectiveness reports freed up hours so she can focus more on her engagements without quality slipping
- Professional-grade work takes less time to produce, and all of it reads and sounds like Honeycomb
- She understands how to use AI responsibly with her sensitive business data, keeping model training turned off and carefully feeding it select information
- By the end of the engagement she was extending the setup on her own, running several self-directed Claude projects
Built with
When Kyle and I started, I had so much I wanted to do with AI but I didn't feel confident I could do it myself. He stood next to me while I built, session after session, and my confidence has genuinely grown, not just in how I use these tools but in knowing the edges of what I can do and when I've got it on my own. I run so much of my practice with these tools myself now, and I trust I can keep going after our work together. I'm very grateful for our time together and the value that this engagement has created for myself and my business by enabling me to work much more confidently and efficiently.