Available for remote work

RAFAELAGUIAR

AI Automation Engineer / Creative AI for Direct Response

I turn copy into finished video and build the pipelines that do it at scale.

See the work → Get in touch

▚bend the flow · click to reveal

SCROLL

/reel01

Selected work.

Avatars, UGC, product and 3D inserts for VSLs and ads.

Character, product and B-roll show up here as one card of six cuts each. The filter opens the scenes one by one.

Program01 / 27

V1 · AVATARS08

● on monitor

AvatarWith the product9:16 · 0:04

● on monitor

AvatarDoctor9:16 · 0:04

● on monitor

AvatarTestimonial9:16 · 0:04

● on monitor

AvatarIn the kitchen9:16 · 0:04

● on monitor

AvatarAt home9:16 · 0:04

● on monitor

AvatarPresenter4:5 · 0:04

● on monitor

AvatarInterview4:5 · 0:04

● on monitor

AvatarIn a blazer4:5 · 0:04

V2 · UGC08

AI-generated video: recipe — **UGC**Recipe9:16 · 0:04

AI-generated video: workshop testimonial — **UGC**In the workshop4:5 · 0:04

AI-generated video: before and after — **UGC**Before and after9:16 · 0:04

AI-generated video: living room testimonial — **UGC**Living room4:5 · 0:04

AI-generated video: product in hand — **UGC**Product in hand9:16 · 0:04

AI-generated video: street selfie — **UGC**Street selfie9:16 · 0:04

AI-generated video: In a jacket — **UGC**In a jacket4:5 · 0:04

AI-generated video: phone video — **UGC**Phone video4:5 · 0:04

V3 · 3D INSERTS08

AI-generated video: blood vessel — **3D insert**Blood vessel4:5 · 0:04

AI-generated video: neurons — **3D insert**Neurons1:1 · 0:04

AI-generated video: drop in vinegar — **3D insert**Drop in vinegar4:5 · 0:04

AI-generated video: pancreas x-ray — **3D insert**Pancreas x-ray9:16 · 0:04

AI-generated video: thermal scan — **3D insert**Thermal scan9:16 · 0:04

AI-generated video: capsule — **3D insert**Capsule4:5 · 0:04

AI-generated video: Holographic brain — **3D insert**Holographic brain1:1 · 0:04

AI-generated video: Capsule and impulses — **3D insert**Capsule and impulses1:1 · 0:04

V4 · SEQUENCES03

AI-generated video: same presenter, six scenes — **Character**Six scenes4:5 · 6 cuts

AI-generated video: six bottles, six settings — **Product**Six settings4:5 · 6 cuts

AI-generated video: a recipe in six cuts — **B-roll**Recipe in six cuts9:16 · 6 cuts

V5 · CHARACTER06

V6 · PRODUCT08

AI-generated video: Neuro Prime and Focus Max — **Product**Neuro Prime and Focus Max1:1 · 0:04

AI-generated video: Nerve Reset and GlycoCare — **Product**Nerve Reset and GlycoCare1:1 · 0:04

V7 · B-ROLL06

1500+

creatives shipped

AI models in production

internal tools built

years in performance

/work02

Tasks I refused to do twice.

LoopVideo

Windows v2.1.0

One photo becomes 2:30 of the person talking, looping with no visible cut.

Generates the clips on Magnific and stitches them seamlessly, ready for a VSL background.
You approve the first video before the rest of the credits are spent.
Three more tabs: images, visual proof and product mockups.

Download for Windows GitHub ↗

the photo one audio per video queued generation step by step downloads the MP4 failed? retry

Avatarize

Windows v1.2.0

A photo and an audio file become a talking avatar video, through your own HeyGen account.

One video per audio file, generated in a queue and downloaded for you.
Signs in through HeyGen's official login. Your password never touches the app.
Trims the silences out of the audio before sending.

Download for Windows GitHub ↗

Also built

Channel pipeline

From a topic, it writes the script, generates narration and visuals, and delivers an MP4 or a CapCut project I finish by hand.

Python · n8n · FastAPI · FFmpegprivate code

topic
script
narration
scenes
MP4

capcut-silence-trim

Finds every pause in the narration and turns it into a cut point inside the CapCut project. No frame is re-encoded, and the timeline shortens in under a second.

Python · FFmpeg · CapCutprivate code

narration
pauses
cuts
clean timeline

browser-session-recorder

A Chrome extension that records what an operator does in the browser, click by click, and outputs a file Playwright replays on its own.

JavaScript · Chrome MV3 · Playwrightprivate code

clicks
JSONL
Playwright
replay

/stack03

What I do.

Generate

AI video, voice, image and avatars, in batches.

HeyGen
Kling
MiniMax
ElevenLabs
Magnific
Higgsfield
Claude

Integrate

I connect the tools over APIs and orchestrate them in n8n.

n8n
Playwright
REST
Webhooks
MCP

Build

When the tool doesn't exist, I write it.

Python
FastAPI
JavaScript
FFmpeg
Whisper

/career04

Career.

since 2024

Senior AI Creator

I turn the copy team's brief into AI video, voice, image and avatars for VSLs and ads.
I build the generation workflows and the internal tools the team uses: a Chrome extension, CLIs and scripts.

Environment: HeyGen · Kling · MiniMax · ElevenLabs · Magnific · Higgsfield · Premiere Pro · CapCut · n8n

Direct Response · Nutra

alongside

Automated YouTube channel

From a topic, it writes the script, generates narration and visuals, and delivers an MP4 or a CapCut project I finish by hand.
Python and self-hosted n8n, with a FastAPI panel. Target cost of about US$1 per video.

Environment: Python · FastAPI · n8n · FFmpeg · Whisper · Playwright · Claude API

Personal project

2022 to 2024

VSL, ads & creative editor

VSL and performance creative editing for the international market.
The first automations: file naming, batch generation and bulk download.

Environment: Premiere Pro · After Effects · CapCut · Photoshop

Freelance

before

Technical IT program, web design and architecture

The technical base came from the program. The eye for composition came from architecture.

Background

/contact05

AI video and the automations behind it. Let's talk about your project.

[email protected]

Remote, Brazil time zone. Native Portuguese, intermediate to advanced English.

GitHub ↗ LinkedIn ↗ Instagram ↗ Download CV →