Video intelligence report

Mommy's Little Sous Chef: Little Sleepies Morning

A mom shows off her baby's Little Sleepies zip-up pajamas, then takes him along to pull her morning espresso shot as her self-declared sous chef.

DcTlocbzf8W_01.mp40:3117 scenes109 words spokenRead as a story

The voiceover is a mom narrating her cozy morning routine with her baby and praising the pajamas rather than teaching steps or arguing a point, so what happens matters more than what's said.

↓ Download PDF

Summary

A mom holding her infant in a matching-toned lounge set introduces her new, chaotic-but-cozy morning routine and praises Little Sleepies pajamas. Close-ups show her working the two-way zipper on the baby's baking-print PJs, highlighting easy diaper changes, the soft fabric, and weekly new prints. She then carries him to the espresso machine, where he watches and reaches for the buttons while she grinds, tamps, and pours. The video ends with her holding him and a striped mug, calling him mommy's little barista and ultimate sous chef.

Tone

Warm, playful, lightly sponsored lifestyle charm

Storyline

    Scenes

    Full transcript

    Since becoming a mom, my mornings have become way more chaotic, but also way cozier. We love little sleepies because they're the coziest, most softest pajamas ever. The two-way zipper makes changing diapers a breeze, and the patterns are just 10 out of 10. They have new designs launched weekly, so there's never a dull moment in selecting our favorite PJs. My son loves to help me make my coffee now, and it's one of my favorite parts of the day, and he loves to push all the buttons and listen to the noises. Soon enough, he'll be mommy's little barista, but right now, he's the ultimate sous chef.

    How this report was made
    1. Scenes. ffmpeg scores every frame for visual change. Cuts become scene boundaries, footage without cuts is sampled at an even interval, and one keyframe is taken from the middle of each scene.
    2. Speech. Whisper transcribes the audio with word-level timestamps, and each word is assigned to the scene it was spoken in.
    3. Lens. A short call reads the transcript and decides what kind of report to write: a story, a how-to (which adds steps), or a talk (which adds key points). This video was read as a story.
    4. Pass 1: storyline. claude-opus-5 reads the contact sheet below (every keyframe with its time range and spoken words) and writes the title, summary, and beats.
    5. Pass 2: keyframes. Keyframes go back to the model in batches of 4, each stamped with its scene number, with the storyline as context. Every caption is checked against the scene it claims, and any skipped scene is requested again.
    Contact sheet: every keyframe with its time range and spoken words