CATELLECT RESEARCH

Understanding cats starts with better evidence.

We build feline-specific AI for behavior understanding, multimodal sensing, and edge intelligence.

Read the latest paperGitHub
Observable evidence04 primary signals
A standing tabby cat showing its overall body posture
Posture01

Body position

An orange Maine Coon walking with its long plumed tail held upright in a natural curve
Tail02

Shape & movement

A young Turkish Angora cat looking out a window with both ears flattened sideways
Ears03

Angle & orientation

A seal-point Siamese cat in natural window light with clear eyes, nose, mouth, and whiskers
Face04

Visible expression

OUR APPROACH

Evidence first. Interpretation second.

We start with what can be observed before inferring what it may mean.

  1. 01Posture
  2. 02Ears
  3. 03Tail
  4. 04Face
  5. 05Interaction
  6. 06Context

STAGE 1

Structured behavior understanding

Observe and organize what is visibly present.

STAGE 2

Evidence-supported interpretation

Interpret only after the evidence has been assembled.

EDGE INTELLIGENCE

From visible signal to local understanding.

A compact evidence pipeline designed to stay responsive and oriented toward edge deployment.

01

Sense

Observe visible signals across the cat and the surrounding scene.

02

Structure

Organize posture, tail, ears, face, interaction, and context as evidence.

03

Catellect-VL-2B

Run compact vision-language reasoning for structured feline understanding.

04

Schema

Return evidence in a veterinarian-designed behavior schema.

LATEST RESEARCH

Preprint · 2026

CAT-VL / 001

RESEARCH → PRODUCT

Research that becomes product.

Our research informs how Catellect observes behavior, combines evidence, and notices meaningful changes in everyday life.

Explore Catellect
01

RESEARCH

Feline behavior understanding

PRODUCT

Helps Catellect interpret what your cat is doing

02

RESEARCH

Multimodal evidence

PRODUCT

Combines signals across the cat and the home

03

RESEARCH

Edge intelligence

PRODUCT

Supports local, responsive visual understanding

Publications

01

Catellect-VL-2B: A Vision-Language Model for Edge-Based Feline Behavior Understanding

2026

arXiv