Advexai offers Advex Composer, an on‑device generative AI vision platform that automates visual inspection for manufacturers. Using a language‑driven UI, operators can define defect‑detection tasks without AI expertise, and the system provides multi‑class classification and pixel‑level segmentation to locate defects in real time, all offline for secure, low‑latency operation.
Funding
$3.5M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
3OFounders
Product
Problem
Manufacturers often rely on manual visual inspection to detect defects such as porosity in welds, dents, scratches, or missing components. This process is labor‑intensive, error‑prone, and limited by the need for skilled operators and on‑site connectivity.
Solution
Advex Composer delivers an on‑device generative‑AI vision system that automates visual inspection in minutes without requiring AI expertise. Users describe inspection tasks in natural language, and the platform configures a model that performs multi‑class classification and semantic segmentation to identify defect type and location. The solution runs entirely offline, eliminating Wi‑Fi dependencies and protecting proprietary production data. Deployments scale from a single‑camera pilot to multi‑camera enterprise setups, enabling rapid rollout across lines and plants. By providing instant defect detection, Advex Composer reduces labor costs, scrap, and downtime while improving inspection accuracy.
Target Audience
Primary customers are manufacturing firms and quality‑control teams that need fast, accurate visual inspection for metal, fabric, leather, and assembly line products, ranging from single‑line pilots to global production networks.
Features
- Language‑driven UI that lets operators configure inspection tasks without coding or AI knowledge
- On‑device inference with zero reliance on internet connectivity for data security and low latency
- Multi‑class classification combined with pixel‑level semantic segmentation to identify defect type and exact location
- Quick‑start hardware bundles (1‑cam pilot, 4‑cam starter) with pre‑installed models for fast deployment
- Scalable architecture supporting multiple cameras and custom models for diverse defect types and product lines
- Real‑time results integration with existing manufacturing execution systems via standard APIs