Skip to content
refs
REF-0295Source onlineBlueprint rev 01

Neo's Agentic ML Engineer Launch Film

A four-minute product launch that moves from a visual labor comparison into a real technical demo, benchmark, and animated system map.

Maker @withneoFor Neo AIOriginal post ↗
Cut timeline · 15 shots03:52
00:0000:3001:0001:3002:0002:3003:0003:30

Field notes / why it works

The many-versus-one tableau becomes visible by 0:02.6 and turns the productivity claim into an instant picture. From 0:42 to 2:17, the film shows task entry, data work, code, a report, and benchmark bars; from 2:17 it explains the system with 11 updates inside one long architecture shot before showing human review at 3:14.

01

Plate 01 / The format in one breath

A widescreen B2B product announcement opens with a simple before/after comparison, puts a calm technical presenter on camera, then spends most of its time proving a demanding workflow inside the product. Once the result has been shown, it widens back out to benchmark evidence, an animated system diagram, human oversight and a waitlist invitation.

Why it works

  • The opening makes the problem visual by 0:02.6: three workers at a long table versus one person with a laptop, before the spoken claim starts at 0:02.1.
  • The promised benefit gets a worked example from 0:42 to 2:17. Screen captures show task entry, data, code, a written comparison and a benchmark instead of relying solely on presenter claims.
  • The architecture arrives after proof, at 2:17. A long animated diagram keeps explaining the system through 11 internal picture changes, then the film lands on human approval at 3:14 and a direct waitlist CTA at 3:39.
02

Plate 02 / Format card

Reference
Platform · aspect · lengthX · 1280 × 720, 16:9 · 3:51.6 · 29.97 fps
Pace15 shots, 14 measured cuts (3.6/min; 3.6–5.4/min if uncertain events are cuts); 42 changes inside shots; 14.5 total picture changes/min; mean shot 15.44 s, median 10.33 s
Script543 spoken words; 149 wpm over runtime, about 179 wpm while speaking; 33 sentences averaging 16.5 words; one question
VoiceLow-register adult presenter, measured and assured, with technical emphasis and mostly falling endings
On screenBy approximate runtime: 51% designed diagrams/cards, 33% screen demo, 13% presenter, 3% computing b-roll; static cover-style opening and black end card included with graphics
CaptionsNo running subtitles. Small diagram labels, phase cards, benchmark text and occasional lower-third identification carry the key nouns.
SoundTonal music-like bed about 12 dB below speech; no clear recurring cut SFX; -18.1 LUFS measured integrated
StructureSix broad beats: problem/contrast → introduction → live task → result → architecture/control → aspiration/CTA. Worked proof starts 0:42; benchmark 2:10; CTA 3:39.

Pro / 13 plates + recipe

The first plates are open. The full dissection is Pro.

Get the timed shots, type and sound specs, and the runnable recipe.

Open the full blueprint
  1. Plate 03 / The hook, frame by frame (first 3 seconds)Pro
  2. Plate 04 / Structure (beat sheet)Pro
  3. Plate 05 / Shot grammarPro
  4. Plate 06 / Visual style guidePro
  5. Plate 07 / Voice and deliveryPro
  6. Plate 08 / Sound designPro
  7. Plate 09 / PackagingPro
  8. Plate 10 / Script templatePro
  9. Plate 11 / Production planPro
  10. Plate 12 / PromptsPro
  11. Plate 13 / Edit recipePro
  12. Plate 14 / QA checklistPro
  13. Plate 15 / Make it yoursPro
  14. Recipe / runnable promptPro