build
Lever
Find where AI actually fits in a business, and the roadmap to make it happen.
the problem
Where does AI actually fit in my business? Small business owners are asking, and the answers are scattered.
why it hurts
Most AI-adoption tools start a step later, at build versus buy or which vendor, so they answer a question the owner has not reached, and nothing gets started.
what i built
Lever, a free diagnostic: thirteen questions that rewrite themselves around the business, a match against 61 use cases, then five to eight opportunities ranked by fit, each with a brief, a four-phase roadmap and a return calculator. I designed and built it end to end.
how it works
Start at the question owners can actually ask.
whyA tool that opens at build versus buy answers a question its buyer has not reached, and trust in this market starts earlier.
Rank by fit, and name what would sink each idea.
whyThe answer has to be usable, so every opportunity carries a complexity rating and a return range.
Let the numbers make the case.
whyThe free tool surfaces opportunities, and the return calculator creates urgency.
in detail
01 / 05
A short structured assessment, then five to eight opportunities ranked for one specific business. Beside it, a sample of the output for an illustrative vet practice.
02 / 05
Question ten of thirteen. An earlier answer said paperwork was a drain, so this one asks which document work hurts most.
03 / 05
Six opportunities for that practice, ranked by fit, each opening a deep-dive memo that uses the owner's own words where it can.
04 / 05
Each deep dive ends in a roadmap: discovery, build, pilot and production, with an owner and exit criteria for every phase, beside the year-one cost and return.
05 / 05
The return calculator. Change any assumption and the annual saving recalculates.
why it's useful
An owner leaves knowing where to start, what it could return and who owns each phase, written so an operator can act on it. Live and free to use.
technical specs
- A short question set maps a business onto a taxonomy of AI use cases. The options rewrite themselves around the business described.
- Each brief streams back on click as it generates.
- Before that, stripping unread fields from the taxonomy payload and adding prompt caching cut input by roughly two thirds, and it still timed out under real load.
live site
Visit the live site↗