I'm Richard. I built solobuilder.ai because I got tired of AI tool reviews that were basically affiliate dumps with the verdict pre-decided by whichever vendor sponsored the post. I needed honest answers to small questions — “is Cursor worth $20/month for one developer,” “should I pay for Windsurf instead,” “what does Supabase actually cost at scale” — and the answers I was finding online weren't research. They were opinions in a trench coat.
So I built APEX, a research engine that does the work I wanted someone else to do. Every article on this site goes through it. Every claim is cited. Every comparison runs across 35+ sources. The reviews are written for solo budgets and one-person teams, not enterprise buying committees.
That's the whole site.
Every article starts with a falsifiable claim. "Is Cursor worth $20/month for a solo dev shipping a B2B SaaS?" Not "Cursor: a comprehensive review." The question constrains everything that follows.
APEX queries the open web through Firecrawl, prioritizing pricing pages, official docs, GitHub issues, changelogs, and benchmark repos. Each source gets scored on authority (gov/official > established media > blogs > forums) and recency (the last 90 days weighs more than 2-year-old reviews).
APEX writes a synthesis. Then it runs an 8-vector adversarial critique against its own draft — checking for overclaiming, weak evidence, missing counterarguments, verdict-evidence mismatch. It argues the bull case and the bear case for the verdict. It grades the whole thing on a 12-criterion rubric. Sections scoring below 7.0 get auto-revised before I see the draft.
I read the APEX draft, cut filler, add nuance, fix tone, and put my name on the byline. Citations stay visible in the final article. Inferences are tagged [INFERENCE — confidence: H/M/L]. Estimates are tagged [ESTIMATE — verify: URL]. Nothing speculative ships unmarked.
APEX is open about its limits. It can't review tools that don't have public documentation. It can't replace a developer who's actually used a tool in production. It can't predict whether a startup will be around in 12 months. What it can do is read more sources than I could in a week, score them honestly, write a draft, and tell me where the draft is wrong before I do.
The byline is a person. The rigor is a system. Both are required.
A flat-fee research service for small teams who want a clear answer to one question: “which AI tools should we actually use, and which are noise?”
You send me your stack, your team size, and the bottleneck you're trying to solve. APEX runs the research — 35+ cited sources, tool-by-tool scoring against your specific constraints, build-or-buy recommendations. I edit the draft, add nuance, and ship a one-page report.
5 business days. $497 flat. No retainer. No upsell to a $5K/month engagement at the end. Currently in beta — taking a small number of clients to refine the deliverable before opening it publicly.
Reserve a beta slot →