We built Monutus because
we were tired of blind cameras
Our founders spent years in operations and engineering watching organizations pay for extensive camera infrastructure — and then fail to prevent the exact incidents the cameras were meant to catch, because the footage only existed for forensics, never for prevention.
The founding moment
In 2021, Arjun Nair was troubleshooting a client complaint at a large manufacturing facility. A worker had sustained a hand injury near a piece of machinery. The plant had 140 cameras. The incident investigation required reviewing 18 hours of footage across six cameras — manually, by a human being — before anyone understood what had happened.
The camera had recorded everything. The AI had recorded nothing.
That same week, he called Dr. Shreya Acharya, who had been doing computer vision research for autonomous vehicles and was convinced the same detection technology being used to make cars safe could be applied to stationary infrastructure. They founded Monutus six months later.
The first version was a single-model PPE detector running on a Raspberry Pi. It was terrible. The model trained on the first real manufacturing customer's floor was significantly better. The third iteration, trained across 12 sites and 500,000 labeled frames, was the product that convinced the first enterprise customers.
Today, Monutus monitors over 50,000 cameras across retail, manufacturing, healthcare, logistics, and smart city deployments in India, the Middle East, and the UK.
How we think
Latency is a moral issue in safety contexts
When someone's life or safety is at risk, an alert that arrives in 5 seconds is worth far less than one that arrives in 0.7 seconds. We're obsessive about latency because our customers' safety culture depends on it.
Privacy-preserving by design, not by disclaimer
We built face blurring, data minimization, on-premise deployment, and role-based access controls into the platform architecture — not into a legal footnote. You should be able to protect people in your facility without exposing them.
Honest about what AI can and cannot do
AI video monitoring is not infallible. Our accuracy numbers are real. Our failure modes are documented. When a model doesn't work in an edge condition, we say so — rather than quietly adjusting thresholds after sale.
The outcome is the product, not the software
Shipping Monutus to a customer is the beginning of our job, not the end. Our success is measured in PPE compliance rates, shrinkage reduction percentages, and incident-free quarters — not license renewals.
The team
The people you'll work with.
Arjun Nair
Former Head of Engineering at a Tier-1 Indian CCTV manufacturer. Spent eight years watching surveillance systems fail to prevent the incidents they were installed to catch. Founded Monutus to fix the intelligence gap.
Dr. Shreya Acharya
PhD in Computer Vision, IIT Bombay. Previously led vision research at a global autonomous vehicle company. Holds 4 patents in real-time video analytics. Designed the model architecture behind Monutus's sub-second detection pipeline.
Vikram Sethi
Ran field operations for a large Indian retail chain before crossing to the vendor side. The person who talks to every new Monutus customer in their first 90 days. Responsible for the outcome-based deployment methodology.
Ananya Krishnamurthy
Product lead for the last three generations of the Monutus platform. Former product manager at a security operations center software company. Obsessed with reducing the time from detection to action.
Where we operate
We're hiring
We're looking for computer vision engineers, enterprise sales professionals, and implementation specialists. If you want to work on technology that actually prevents harm and improves the way organizations operate, we'd like to talk.
View Open Roles