Building mxstermind in Public — Entry 01

Most agencies hide the messy middle. mxstermind does the opposite — not because transparency is fashionable, but because our buyers are technical operators who will verify everything anyway. They will open Discord, click live URLs, read Lighthouse scores, and ask for escrow milestones before they trust a deck.

Publishing in public is a filter. It attracts founders who want accountability and repels buyers who need a lowest-bid vendor. That is intentional. We cap at four clients per quarter; the cost of a bad fit is higher than the cost of saying no on intake.

This series is not a growth hack. It is a record of how we think, what we ship, and what we are negotiating — written the way I would explain it to a peer, not a marketing department.

The portfolio on mxstermind.com is not a greatest-hits reel. It is a curated set of high-complexity deliveries: CarSpotLive shipped to the App Store as a full mobile product — maps, Firebase, community flows — not a landing page dressed as an app. Passle is a creator monetization platform with subscriptions, pay-per-view, tipping, and a Keys wallet, live on an invite-only launch. Drain.cx and DirectFiber show the other end of the spectrum: Figma-to-production commerce and enterprise billing portals where UX and engineering cannot drift.

We also show archived work without embarrassment. Cascade Markets, the SUI Blockchain App (BAM), Jarro AI, Lava.pw, and the LinkedIn automation platform are proof that we have operated under hard deadlines, fixed budgets, and discontinued clients. Client discontinued is not a failure of delivery — it is a fact of venture outcomes. The studio's job is to ship what was agreed; what happens after handoff is not ours to control.

If you want depth on a specific delivery, start with the case studies for CarSpotLive (/portfolio/carspotlive) and Drain.cx (/portfolio/drain-cx), then read how we rebuilt a Solana-era codebase for SUI in fourteen days on the blog. For how we operate, read /process and /developers — those three references tell you more than any capabilities slide.

Two engagements are visible at category level only. The AI voice receptionist prototype targets dental clinics in France — appointment booking and patient queries handled by voice, with no human operator required. We are in active development and deployment discussions; clinic names and commercial terms stay confidential.

Crystal Wars is a Unity strategy game — civilization selection, troop and king mechanics, ranked play. Public information is limited to category, engine, domain (crystal-war.com), and our role in scoping design through multiplayer infrastructure. Budget, client identity, and milestone structure are under NDA. We will not leak scope to win attention.

A third slot is reserved for confidential engagements. If you are reading this and need NDA-only work, that is normal — start on Discord with your constraints listed plainly.

Outcome-based does not mean vague promises. It means milestones with deliverables, acceptance criteria, and a named signatory on your side. We map those milestones to escrow-friendly payments, weekly updates in Discord, and written change requests when scope shifts. The process page on this site is the operational version of that philosophy — discovery before pixels, sign-off before build, staged delivery instead of a big-bang reveal.

The developer platform pages exist for the same reason: buyers want to know how we think about stack choices, AI systems, blockchain delivery, and automation before they commit. We document what we actually ship — Next.js, TypeScript, Cloudflare, Firebase, SUI, Stripe, n8n — with reasoning, not logos.

Build-in-public extends that honesty to narrative. When we miss a target, we say so. When a client discontinues, we keep the case study as proof of execution. When we are full for the quarter, we tell you directly.

If you are established, have real budget, and care about perception matching product quality, this series is for you. Come to intake with a link, an outcome, a timeline, and a budget range. Use Telegram with the pre-filled message or open Discord — we reply same-day when the brief is serious.

Do not ask us to compete on price with template shops. Do not ask for unlimited revisions without change control. Do ask for references, live URLs, and a milestone map — we will provide them.

Entry 01 ends here. The site will keep evolving in public: portfolio visuals, performance work, and new case studies as deliveries complete. If you want the next entry to cover a topic — escrow mechanics, mobile App Store war stories, or how we scope AI voice — say so in Discord. We read it.