oof! finally able to share the final piece!!

if you haven't been keeping up with my updates:

⭐ on day 1 I brainstormed alongside Gemini, merging my ideas with theirs and ending up with an outline of what I wanted to do for this challenge

⭐ day 2 I collected the raw footage I filmed and started putting together a folder with all my assets

⭐ day 3 I dove into Envato for the first time in years and found stock video, music and sfx, and used AI to alter both one of my own clips and a stock one

⭐ day 4 was for the final touches, where I recorded the voiceover and put all the other assets together into the video you can see below - my final submission!

(I also attached a screenshot of what my final project folder looked like - organized by source of each asset)

my biggest AI win was definitely the clip where I asked it to replace my water bottle with the coffee machine and it mimicked my rings and tattoos without me even asking

my AI flop, where I decided to intervene because AI just didn't do it justice, was the voiceover - it just felt off and emotionless, and instead of trying to tweak it to perfection I decided to just pick up the mic and do it myself!

I'm also gonna be sharing this final video on my Instagram @freelancingwithsilvia - feel free to follow me there for all the freelancing vibes 🤩

I love it, Sílvia! I watch your YouTube videos too. 🙂‍↔️

AI Spec Ad This spot was created entirely using AI tools, from start to finish — no cameras, no actors, no shoot.

It began with a base image generated in ComfyUI, building out the visual concept of "Sleep Analysis" — a human face transformed into a data dashboard, with sleep metrics, heart rate, and apnea alerts layered directly onto the skin, as if technology could read the body in real time.

From that image, the motion and narrative were built in Seedance, animating each interaction: the tap on "Sleep Score" that triggers a purple-and-orange gradient of light, and the cut to "Sleep Apnea," where everything goes dark and a scanning light sweeps across the face as she sleeps.

The result is an exercise in AI-driven creative direction — showing how a product campaign, in this case built around the concept of wearable-free sleep tracking, can be fully constructed using generative models, from the first frame to the final animated piece, while maintaining visual consistency, cinematic lighting, and a clear narrative in just 15 seconds.

No humans were filmed. No locations were rented. 100% AI, from image to video.

This looks fantastic