Industrial Vision AI & Robotics Platform

Compose vision.
Create intelligence.
Act in the real world.

GX-Arc turns computer vision from a coding problem into a building problem — drag together a pipeline, or simply describe the task, and ship hardware-optimized production code. Then put that intelligence into robots and machines that do the work.

Identity

Two letters. One discipline.

GX is the engine — the way a pipeline is composed, extended and run anywhere. Arc is the intelligence that composes it.

G
01 Generative — pipelines assembled, not hand-written
02 Genius — models that reason about what they see
03 Generational — built to outlast one model cycle
X
01 Extensible — every node is a swappable part
02 Execute — from canvas straight to production code
03 Evolve — pipelines that improve as models do
Arc — your intelligent AI identity. This defines the intelligence that drives every pipeline and powers every action in the real world.
Products

Compose. Create. Act.

Build a pipeline by hand, or describe the task and let Arc build it for you. Then put that intelligence into machines that do the work.

01 / STUDIO

GX-Studio

Compose · Human → AI Pipeline

A drag-and-drop canvas for building IVA and computer-vision pipelines node by node.

Sees
Every node on the canvas — inputs, models, transformsdrag / connect / configure
Understands
How data shape and requirements flow between nodestype-checked connections
Does
Dry-runs the graph, then exports optimized codePython / C++ · platform-aware
Open GX-Studio →
02 / AGENT

GX-Agent

Create · AI Agent → AI Pipeline

Describe a task in natural language. Arc assembles the same class of pipeline for you.

Sees
The task as written — intent, constraints, target platformplain-language query
Understands
Which nodes and models satisfy the requestmaps intent → graph
Does
Builds the pipeline on canvas, ready to inspect and runeditable in GX-Studio
Open GX-Agent →
03 / ROBO

GX-Robo

Act · AI → Physical World

Robotics — machines doing work. GX-Robo puts the intelligence into machines.

Sees
The physical scene through the machine's cameraslive vision pipeline
Understands
What the pipeline's output means for the task at handperception → decision
Does
Drives the machine to act on it in the real worlddecision → motion
Open GX-Robo →
GXP_01 · built in GX-Studio
GX-Studio canvas showing an object detection pipeline
From canvas to hardware

One pipeline, four stages.

What you build in the UI is what ships — inspected, tested, and compiled for the exact hardware it will run on.

01

Build

Compose the pipeline visually in GX-Studio, or generate it with GX-Agent.

canvas
→
02

Dry run

Select real inputs and models per node, and run the graph live in the UI.

preview
→
03

Generate

Emit production-ready code in Python or C++ straight from the graph.

codegen
→
04

Deploy

Pick the target platform — the output is compiled and tuned for it.

hardware-aware
Hardware-aware optimization

Same pipeline. Any silicon.

One inference pipeline, optimized for the hardware you already have — from Intel CPUs and NVIDIA GPUs to edge NPUs and heterogeneous CPU+accelerator systems.

Intel CPU · OpenVINO
Run the complete AI pipeline directly on Intel processors using Intel OpenVINO.
CPU-native inference
OpenVINO-optimized execution
No dedicated GPU required
Efficient deployment on Intel edge and server platforms
Ideal for CPU-only environments and cost-sensitive deployments
Best for
Edge gateways, industrial PCs, servers, and systems without dedicated accelerators.
NVIDIA Jetson · GTX · RTX
Accelerate the complete inference pipeline on NVIDIA GPUs for high throughput and parallel execution.
Supports NVIDIA Jetson, GTX, and RTX platforms
GPU-accelerated inference
Parallel inference execution
High throughput for demanding workloads
Suitable for multi-stream and real-time applications
Best for
Computer vision, video analytics, robotics, and high-throughput AI workloads.
Intel CPU + NVIDIA GPU
Distribute workloads intelligently between the CPU and GPU to maximize overall system performance.
Intel CPU + NVIDIA GPU processing
Split pre-processing / inference / post-processing
GPU handles compute-intensive inference
CPU handles data preparation and result processing
Better resource utilization and pipeline concurrency
Best for
High-performance systems where CPU and GPU resources need to work together efficiently.
In development — support for all listed NPU platforms is in the integration pipeline
Rockchip RKNN · NovaTek · Kneron · More
Deploy AI inference directly on dedicated Neural Processing Units (NPUs) built for edge computing.
Supports Rockchip RKNN, NovaTek, Kneron, and other NPU platforms
Dedicated neural acceleration
Low-power AI inference
Hardware-optimized execution
Designed for compact and resource-constrained edge devices
Best for
Embedded devices, smart cameras, IoT, consumer electronics, and battery-powered systems.

Target selected once, per pipeline — compilation, memory layout and kernel choice adapt automatically.

Your application stays consistent while the execution layer adapts to the silicon — CPU, GPU, NPU, or heterogeneous combinations.

Model registry

Every AI node needs a model. We keep 21, ready.

A production registry of generic and use-case-specific models, trained in-house on millions of samples — drop any of them straight into a GX-Studio node, or have our team train the one you actually need.

21
Production models
6
Model categories
7+
AI modelling engineers
1M+
Training samples / model
In-house detection models
Person
Vehicle — car, truck, bus, motorcycle, bicycle
Face & Hand
Road Signs & Traffic Signals
Tablet & Capsule
PPE — vest, gloves, safety shoes, goggles, helmet
Fire & Smoke
License Plate
Tools & Equipment
Barcode & QR Code
Best for
Detect the objects, people, assets, and events that matter to your application.
In-house segmentation models
Person
Vehicle
Road Signs & Traffic Signals
Tablet & Capsule
Industrial Components
Road & Lane Areas
Fire & Smoke
Best for
Understand exactly where an object or region exists — down to the pixel.
In-house classification models
Face — Mask / No Mask
Defect / No Defect
Damaged / Undamaged
Good / Reject
PPE Compliance
Tablet / Capsule Type
Road Sign Type
Best for
Turn visual observations into actionable labels, grades, and decisions.
In-house keypoint models
Human Pose
Face Landmarks
Hand Landmarks
Object Keypoints
Industrial Part Alignment
Best for
Go beyond detection with precise structural understanding.
In-house OCR models
License Plate Number
Road Signs & Boards
Product Labels
Serial Numbers
Batch / Lot Numbers
Expiry Dates
Manufacturing Dates
Barcodes
QR Codes
Meter / Gauge Readings
Industrial Markings
Packaging Text
Best for
Convert real-world text into structured, machine-readable data.
In-house feature extraction models
Person Re-ID
Face Embeddings
Hand Embeddings
Vehicle Embeddings
Tablet & Capsule Embeddings
Generic Object Embeddings
Visual Similarity Embeddings
Best for
Transform visual content into embeddings for search, matching, retrieval, and re-identification.
Beyond the registry

No model fits? Our modelling team builds it.

Data engineering, architecture design, and training run by an in-house team of 7+ — for the exact use case, not the closest generic match.

Discuss custom training →
Get started

Bring a vision task. Leave with a pipeline.