Before any client, we test at home. New tool, new prompt, new workflow: it runs in our daily work first. What survives becomes a class, a talk, a system, a video.
Research isn't something you hire on its own. It's the ground under the other four practices. When something new is real, we write about it on the blog and newsletter before it becomes an offer.
We run dozens every month and keep the ones that survive real use: Claude, ChatGPT and Gemini for conversation; Claude Code and Cursor for code; Midjourney and Flux for image; Sora, Veo and Kling for video; Suno and ElevenLabs for audio. Each with its place, limits and moment.
conversation · code · image · video · audio · avatar
Ask what we useEvery project we ship started as a test nobody paid for. The Senac grader, Continental's mini-films and the SXSW app were internal experiments before they became contracts.
prototypes · pilots · proofs of concept
See the casesThe public side of research. When something new is real, we write about it: what changes, who it serves and how to use it without hype.
analysis · guides · what to ignore
Follow the blogWhat we post became part of the work. @fepacheco covers productivity, process and tools; @paulofaguiar covers visual creation with AI.
@fepacheco · @paulofaguiar · @cr_ia.pro
Follow on InstagramWe run dozens of tools a month. Half a dozen survive real use.
