Artificial intelligence & multimodal systems
Models and experiments that connect language, image, sound and structured information without treating intelligence as a single medium.
↗Independent research laboratory · Santa Catarina, Brazil
Independent research across artificial intelligence, software, embedded systems and scientific instrumentation.
Research coordinates
THE LABORATORY
Cecil-IA Labs develops experiments at the boundary between artificial intelligence, software, hardware and scientific observation.
We work by making ideas testable: building interfaces, instruments and prototypes that reveal what a concept can — and cannot — do in the real world.
01 RESEARCH FIELD
Models and experiments that connect language, image, sound and structured information without treating intelligence as a single medium.
↗Interactive systems designed around presence, feedback and low-latency exchange between people, models and computational tools.
↗Software as an instrument of investigation: observable workflows, reusable components and environments for disciplined experimentation.
↗Physical prototypes that bring computation into motion, sensing, actuation and the constraints of the material world.
↗Explorations that combine measurement, simulation, optics and computation across disciplinary boundaries.
↗02 SELECTED TRACES
A partial view of public, private and experimental work. It is not a complete catalogue of the laboratory.
Interface record · supplied project materialP–01 AI systems / experimental interfaces
An experimental environment for composing prompts, model resources and agent flows. The work explores how complex AI behavior can remain legible while it is being designed.
Prototype record · supplied project photographP–02 Robotics / embedded systems
A wheeled physical platform used to investigate mechanical articulation, actuation and iterative control. Presented here only as a record of the prototype.
AI-generated conceptual frame · not a documentary recordP–03 Multimodal media studies
Studies of image-to-video and generative visual pipelines as material for experimentation, observation and new forms of technical storytelling.
Public project · package page recordP–04 Public software / React & D3
A maintained, typed React component built on the d3-cloud layout, with a current TypeScript source tree and modern React-compatible tooling.

Public channel record · supplied screenshotP–05 Science communication / exploratory media
An audiovisual channel where scientific ideas, technology and learning are translated into compact exploratory narratives for a broad audience.
03 WORKING PRINCIPLES
We begin with what is not yet understood and keep uncertainty visible throughout the work.
Prototypes are instruments of inquiry: they expose constraints, assumptions and new questions.
Software, hardware and scientific observation are treated as parts of the same investigative field.
Privacy, legibility, accessibility and restraint belong inside the experiment, not after it.
We move between reading, modeling, making and observing. Each cycle is allowed to alter the question.
TECHNOLOGIES & DOMAINS
CECILIA REMAINS PRESENT
For research conversations, interdisciplinary questions and carefully framed experiments.
juninho@cecilialabs.com ↗