MATZU — PLAN. WALK. ASK.
Private homepage chapter revision, 43.9 seconds.

00:00:00.300–00:00:03.580
Matzu. Your AI tour guide.

00:00:04.100–00:00:05.540
Plan your day.

00:00:06.200–00:00:11.440
Matzu, plan me a London day trip
with six iconic sights.

00:00:14.700–00:00:16.740
Walk and listen.

00:00:20.100–00:00:26.650
Welcome to Tower Bridge! Just look at this iconic silhouette,
a true masterpiece of Victorian engineering.

00:00:27.200–00:00:28.880
Ask anything.

00:00:29.200–00:00:33.200
Wait, Matzu — why does Tower Bridge open?

00:00:33.500–00:00:39.260
It opens to allow tall ships to pass through!
Want me to look up when the next opening is scheduled?

00:00:39.650–00:00:43.570
Matzu. The tour guide that talks back.

Visible result caption (11.600–14.700): Your sights and timings, together in one day.
Visible walk caption (16.700–20.100): Hear each story as you arrive.

WHAT THIS EDIT SHOWS
Existing approved demo footage and unchanged voice excerpts, shortened and reordered into Plan, Walk and Ask, preceded by the full-screen original-style opening and three 2-second chapter slides with three wide horizontal bars stacked vertically. The restaurant/fourth chapter is not included. The full demonstration remains available on the press page.

The planning and London field recordings are separate app sessions (six and ten stops respectively), not one continuous trip or a new device test. Planning speech is scripted dubbing. The London GPS approach is time-lapsed at the existing approved 21x speed; arrival returns to normal speed. The detail card is opened manually. The video preserves that disclosure and the original full app screen.

The original field recording is silent; guide narration is recreated from Matzu's production narration prompt. The question and model-generated answer were recorded in a separate native-audio session using Matzu's released base prompt, Short mode and Aoede voice, not inside the phone app. They do not demonstrate device latency, microphone recognition or a completed live lookup. The offered bridge-opening lookup is not shown as performed. Its contextual Tower Bridge image is a founder-owned 2019 photo, not a frame from the 2026 walk.

The full original voice cues are retained without new speech generation. The long itinerary voiceover is omitted entirely; the short result shot uses a visible caption and music instead. No completed booking, automatic card opening or unlimited free use is claimed.

MUSIC CREDIT
Carefree by Kevin MacLeod (incompetech.com).
Licensed under Creative Commons Attribution 4.0:
https://creativecommons.org/licenses/by/4.0/
Source: https://incompetech.com/music/royalty-free/index.html?Search=Search&isrc=USUAN1400037
Edited, volume-adjusted and mixed with the existing narration.

The complete original intro and chapter voice files are reused; the 2.04-second Walk and listen file extends 0.04 seconds past its 2-second title card, retaining its silent tail without accelerating or clipping speech.