Industrial AI

    Connect your machines. See what they're telling you.

    We install cameras and sensors on your production line, connect to your PLCs, and train AI to catch defects, predict failures, and monitor quality in real time.

    >100Projects delivered
    67%Repeat clients
    2018Founded
    ZürichSwitzerland
    Discuss your production challenge
    • Vogelwarte
    • UMAMI
    • Syntegon
    • ETH Zürich
    • GHM
    • DigiHealth
    • EPFL
    • Unil
    • REA
    • VY

    Sound familiar?

    You know something's wrong when scrap rates go up. When the bearing fails. When the customer complains. By then it's too late, and you're left guessing what happened.

    Why Octanis

    Eyes and intelligence where you didn't have them.

    Cameras that see defects operators miss. Sensors that track what matters. AI that spots patterns across months of data, running locally on your factory floor.

    Raw feedAI Vision

    Process

    How we work

    01

    Audit

    We visit your factory. Document your machines, protocols, and what you actually need to monitor.

    02

    Install

    Cameras, sensors, PLC connections. We integrate with what you already have.

    03

    Train AI

    Using your real production data. Defect detection, anomaly detection, predictive maintenance.

    04

    Deploy

    Running on your factory floor. We stay involved until it works.

    Projects

    What we build

    UMAMI
    Custom MES

    UMAMI

    PLC integration, real-time data, and edge AI for microgreens production.

    Syntegon / Makat Candy Technology
    Consulting

    Syntegon / Makat Candy Technology

    AI workshop for confectionery production managers. Computer vision for quality inspection on mogul and starchless lines.

    Aluminium Defect Detection
    Manufacturing

    Aluminium Defect Detection

    AI-powered visual inspection for surface defects on aluminium parts.

    Factory Agentautonomous-loop
    Working

    Trained on your reject pile, not ImageNet

    Generic vision models don't know what a bad gummy bear looks like. Ours do, because we fine-tune them directly on your production line, with your real rejects, your lighting, your conveyor speed.

    The model runs on a local GPU box under your desk. No data leaves the building. When it flags something, your operator gets a live overlay on the HMI. Not an email three hours later.

    Integration

    We speak your machines' language.

    Connect any industrial protocol to a single gateway. Your data stays on-premise, with optional secure cloud forwarding when you need it.

    Common questions

    Let's talk

    Ready to boost your OEE?

    Tell us about your project.

    Sam Sulaimanov

    Sam Sulaimanov

    MSc ETH · Electrical Engineering

    Discuss your project