Lifescape

An evidence-backed retirement decision engine

Decide where retirement still works.

Lifescape turns your non-negotiables, town evidence, and future-life priorities into an explainable shortlist—without hiding failed options or guessing through missing data.

See the finished demo See how it decides

The public demo uses invented evidence. It is for exploring the method, not choosing a home.

Synthetic field 10 towns · 17 metrics
Gates eliminate Weights rank Evidence decides

01 / The difference

A shortlist is only useful when you can see why it exists.

Most comparison tools collapse everything into one score. Lifescape separates the decision into hard gates, ranked preferences, source quality, and sensitivity—so a town cannot compensate for failing something essential.

Hard gates Non-negotiables can eliminate a town.

Budget, critical access, and missing essential evidence stay decisive.

Weighted ranking Only eligible towns compete on preferences.

Comparable evidence is normalized before priorities shape the ranking.

Visible uncertainty A rank is tested, not presented as destiny.

Sensitivity runs show whether the leading group is stable or fragile.

02 / How it works

From “maybe there” to a decision you can inspect.

See the finished decision →
  1. 01

    Set the decision frame

    Define the purchase ceiling, planning age, household, and the towns worth comparing. These inputs remain visible and editable.

  2. 02

    Inspect the evidence

    Review completeness before scoring. Blank critical fields remain unknown; the engine does not invent a value to keep a candidate alive.

  3. 03

    Read ranked and blocked outcomes

    See the eligible ranking, the criteria behind each result, stability across 1,000 simulations, and every town that failed or could not clear a gate.

03 / The output

The answer is not just a winner.

It is a decision record: what qualified, what failed, which criteria mattered, how stable the result was, and where the evidence still needs work.

Synthetic example Williamsburg leads this field.
Eligible
Ranked with criterion detail
Blocked
Still shown with gate reasons
Uncertain
Missing evidence remains explicit
Local app
Reports and SQLite provenance

04 / Local by design

Use the web demo to learn it. Run locally for your real work.

The hosted site accepts no inputs: it explains the method and shows a finished synthetic example. The installable app lets you import your own CSV, processes it on your computer, and produces downloadable reports plus SQLite provenance.

uv sync --locked --extra dev uv run retire app
View the source on GitHub ↗

Start with the method

See what changes when failed options stay visible.

See the finished demo