Agentryx AI

Itineva

Travel plans built like websites: versioned, shared, alive.

Itineva turns destination, dates, travelers, and taste into a shareable trip page. A staged generator drafts a skeleton, fills day-by-day events, enriches places, and checks routes before publish.

Inputs
Destination · dates · travelers · taste
Output
Versioned trip page
Generation
Streamed · 4-stage
Stage
Core candidate
01 — Why a website beats a PDF

Plans change. Pages change with them.

A trip page with versioning keeps every traveler on the same plan. Each version has a stable share link, and the featured version is what the group sees by default.

02 — A four-stage generator

Skeleton, day plans, enrichment, autofix.

The generator streams a skeleton, creates day plans in parallel, enriches places with real-world data, and runs an autofix pass before publish.

03 — Route checks against reality

We refuse to suggest impossible days.

Day plans are checked against transit time, opening hours, and pace. If a day is too tight, the page flags it before publish.

Product proof

Concrete evidence before broad claims.

Each block is grounded in internal research and product implementation facts. Public copy does not overstate launch status, legal posture, or implementation scope.

Generator evidence

Generation is staged, not one-shot.

Repo research identifies Stage A skeleton streaming, Stage B parallel day-event generation, Stage C Google Places enrichment, and Stage D autofix/watchdog before publish.

Share evidence

The shared artifact is versioned.

The implementation uses Project, Run, and Version models, a featured_version_id, publish/feature/share/import flows, public gallery behavior, and shared itinerary attribution.

Reality evidence

Routes and places are part of the product surface.

The product facts identify Google Places-based place enrichment and Google Routes-based route validation, with public copy limited to checks that the repo actually supports.

A trip plan should be a living artifact, not a stale attachment.