Building in public means being honest about what didn't work. These tools were real experiments. Some had the right concept but the wrong implementation. I write them down here because the reasoning matters.
A six-persona AI review panel that evaluated nonprofit fundraising appeals and grant narratives. Each persona stood in for a different donor or funder and returned a score, written feedback, and a revision plan.
Security. The tool required users to paste real donor data, appeal copy, and grant narratives into a shared online tool. That creates real risk around data handling, confidentiality, and organizational liability, and I wasn't comfortable with it once I understood it more deeply.
The concept is sound. If this comes back, it comes back with proper data architecture, no shared inputs, clear security documentation, and explicit guidance on what should and shouldn't go into an AI tool.
A $197, 30-minute check of where a local service business was losing leads: the website form, missed calls, how fast replies went out, and whether anyone followed up.
Free tools can now run that same check in a few minutes. Charging for something you can get free didn't sit right with me, so I stopped selling it.
The part that was always worth paying for is the fixing. Missed call text-back, instant lead replies and quote follow-up are now hats you can pick in Hat Trick.
A done-for-you auto-reply that texted every new lead within a minute, plus a DIY follow-up guide and a quick-start guide for real estate agents.
Each one fixed a single chore. Most owners I talk to are wearing six hats, not one, and a one-trick setup didn't give them enough time back.
The same fixes, folded into a bigger menu. Pick them alongside the other chores eating your week in Hat Trick, or read the free speed to lead guide.
A free tool that generated a six-section draft AI policy in minutes, covering acceptable use, data privacy, and disclosure. Built when generative AI was new to the sector and most orgs needed exactly that: something basic to get started with.
AI capability has grown enormously in the last 6 to 12 months, and it isn't slowing down. What counted as a complete policy a year ago doesn't anymore. This tool never addressed what happens once AI stops waiting to be asked and starts taking action on its own: sending things, scheduling things, and making small decisions without a person checking each time. That gap didn't exist when this tool was built. It exists now, and it's growing.
A framework built for exactly that, including the part this tool was never designed to cover. See the Agentic Governance Framework on the Nonprofits page.