Local AI · IoT · Software · Cloud · São Paulo, Brazil

We build local AI, devices and software. You keep the code.

Plugen builds AI products that run on your own infrastructure, IoT hardware, custom software and cloud for companies that want to control their own technology. At the end of every project you get the source code, the documentation and access to everything we built.

See projects Talk to us about a project
CompanyPlugen
Based inSão Paulo, SP
AreasLocal AI, IoT, software and cloud
DeliverySource code and documentation, owned by you

Projects

What we have built

Each project shows the problem, what we delivered and the result after launch.

Illustration generated by Fabularte: Winston writes in his diary, hidden in the alcove out of the telescreen's sight.
Nineteen Eighty-Four · Chapter 7 · illustration generated by Fabularte
Illustration generated by Fabularte: the forest of white mushrooms from Journey to the Center of the Earth.
Journey to the Center of the Earth · Chapter 27
Product · Local AI · WebIn beta

Fabularte

The classics, illustrated from the first chapter to the last.

14classics illustrated
943illustrations generated
What it is
A platform that reads the whole book, picks the scenes a reader would remember and draws them at the exact point in the story where they happen.
How it works
Every illustration comes from the book's own words: the plate's “X-ray” shows which passage produced each element. The library already includes Dracula, The Odyssey, The Wonderful Wizard of Oz, Dom Casmurro and more.
Next step
Fabularte+: the same character from start to finish, drawn from a single reference portrait. Alongside it, a closed beta for authors to illustrate their own books.
What we built
The whole pipeline, from reading the book to drawing the plate, runs on open-weight models on our own GPUs (CUDA), with no third-party AI APIs.
Visit Fabularte ↗

Services

Four areas, one team

The same team handles the AI model, the device, the software and the cloud, so nothing gets lost between vendors.

Local and applied AI

AI products that run on your infrastructure, with open-weight models and GPU inference. Your data stays in-house and you are not tied to a third-party API. Fabularte was built this way.

  • Open-weight models: language, image and vision
  • Local GPU inference (CUDA)
  • Fine-tuning and evaluation on your own data
  • Text and image generation pipelines

IoT infrastructure

From prototype to production at scale: we design the device, write the firmware, connect it to the cloud and support it after launch.

  • Hardware prototyping
  • Large-scale production
  • Connectivity and cloud deployment
  • Support and maintenance

Software development

Custom systems and legacy modernization, with code your team can maintain afterwards.

  • Custom systems and platforms
  • Legacy modernization
  • Cloud deployment

Technical consulting

Architecture, systems integration, cloud migration and cost control (FinOps).

  • Solution architecture
  • Systems integration
  • Cloud migration and FinOps

How we work

From first call to operation

01

Assessment

We understand the problem, timeline and budget, and send back a written scope.

02

Prototype

We validate the solution at small scale before investing in the final version.

03

Production

We build and deploy, with partial deliveries on the agreed schedule.

04

Operation

We hand over code and documentation, and keep supporting you if you want.

Contact

Tell us what you need to build.

Send a message with the problem and the deadline. We reply within one business day with next steps.

Emailcontato@plugen.io

We use your details only to reply to this message. Privacy policy.