A calculator that turns "you should visit your parents more" into an exact, unavoidable number.

Sacred Death Collective case study thumbnail

Kairos Craeft

## The Challenge

The Sacred Death Collective's core message — sit with mortality instead of flinching from it — is easy to say and hard to make land. It needed a way to make the abstraction of "time is limited" concrete enough that a visitor actually feels it, not just reads it.

## The Build

A standalone calculator, linked from the SDC site, that combines country-level life-expectancy data with how often someone actually visits to produce one blunt number — "X more times to see your parents," or "X more sunsets" if you're asking about yourself — delivered with a deliberate, paced reveal animation rather than an instant result. It also captures an optional six-month check-in reminder by email, turning a one-time gut-punch into a follow-up touchpoint.

## The Result

A shareable, fully on-brand interactive tool that converts SDC's philosophy into a two-minute experience, with lead capture and analytics tracking built into every calculation and every reminder signup.

Try it yourself: <a href="https://www.sacreddeathcollective.com/timewithyourfolks" target="_blank" rel="noopener noreferrer">Time with Your Folks</a>.


Two systems, one client: an AI reading engine that synthesises five self-discovery frameworks into one report, and a guest portal that replaced spreadsheet-chasing for every retreat.

## Inner Map — an AI-synthesised self-discovery reading engine

### The Challenge

Taozen wanted to offer guests a genuinely personal reading — astrology, Human Design, Gene Keys, Enneagram, MBTI — but synthesising five systems into one coherent, on-brand report by hand doesn't scale past a handful of people. Off-the-shelf reading generators produce shallow, template-y output that reads generic and doesn't hold Taozen's voice.

### The Build

An AI reading engine deployed on Vercel and embedded straight into the Taozen Webflow site. One shared, book-grounded knowledge base feeds two routes: an instant on-screen version for browsing the site, and a richer async version delivered by email as a branded PDF, with a hidden "deep mode" that upgrades the model for guests who want more depth. A convergence layer cross-references all five systems so the report reflects real per-person patterns instead of stock zodiac copy.

### The Result

Guests get a personalised reading in minutes on-screen, or a full branded PDF report by email shortly after — all running off one knowledge base kept cheap to run through prompt caching and selective content injection, so cost per reading stayed low as the system grew deeper across five-plus shipped releases.

Try it live: <a href="https://www.taozenfamily.com/inner-map?im=deep" target="_blank" rel="noopener noreferrer">Taozen's Inner Map</a>.

## Retreat Onboarding Dashboard

### The Challenge

Every retreat meant manually chasing guests for signatures, health forms, and payments across spreadsheets and email, while mentally tracking tiered pricing, early-bird and ad-hoc discounts, returning-member exemptions, and multi-installment plans by hand, per guest, per retreat.

### The Build

A guest portal and admin dashboard built on Next.js and Supabase. A single domain engine encodes every pricing and checklist rule — deposit/balance splits, room pricing, discounts, unique payment references, returning-member skips, installment plans — so each guest's checklist and payment amounts generate themselves correctly. The admin board gives the team one guest-by-step grid across every retreat with overdue flags, while each guest gets their own portal showing live progress and exactly what they owe.

### The Result

The onboarding logic that used to live in someone's head now runs as one accountable system, with real guest data behind proper access controls — no client-side key can read the tables, everything goes through a server-only service role with page-level allow-listing. The core engine and dashboards are live in production; automated emails, payment auto-matching, and e-signing are the next phases, already shaped into the data model.
