Ailytics bags US$7.4M Series A to make old CCTV cameras earn their keep

[L-R] Ailytics co-founders Prateek Manocha (CTO) and Lenard Tan (CEO)

Most industrial CCTV cameras lead a thankless life. They record everything, are watched by almost no one, and only become important after something has gone wrong, when their footage turns into evidence rather than prevention.

Singapore-based Ailytics has built a business on reversing that order. The vision AI startup has raised US$7.4 million in an oversubscribed Series A round led by Tin Men Capital. SG Growth Capital and Japanese cloud camera firm Safie also took part, along with one of Japan’s largest construction companies (not named), family offices and returning investors.

Also Read: How Groundup AI redefines what smart factories expect from their machines

The round comes two and a half years after Tin Men led Ailytics’s US$2.7 million pre-Series A in April 2024. In that time the startup says it has grown from 70 projects in four countries to more than 400 deployments in 15, including the US, Australia, the Netherlands, Hong Kong, Japan, the UAE and Saudi Arabia.

Software for cameras that already exist

Ailytics doesn’t sell hardware. It adds a software layer to the cameras already bolted onto scaffolding, gantries and warehouse ceilings. Its platform, Ailyssa, works out three-dimensional spatial relationships from a single flat video feed. In practice, that means telling the difference between a worker who only looks close to an unguarded edge and one who actually is.

In June, the company launched AilyssaX, a contextual reasoning engine that checks whether multi-step procedures are followed in the right order, not just whether someone is wearing a helmet. Ailytics says five years of live deployments have trained it to cope with rain, dust, poor lighting, blocked views and cameras older than some of its engineers.

The headline numbers are striking. Customers report a 73 per cent drop in their total recordable incident rate (TRIR), the standard industry measure of work-related injuries per hours worked, and 84 per cent fewer safety non-compliances. The caveat: these figures are customer-reported, and the company has not disclosed an independent audit or the baselines behind them.

The problem itself is not in doubt. Ailytics cites estimates that nearly three million workers die each year from work-related accidents and diseases, with construction, manufacturing and marine accounting for 63 per cent of fatal occupational injuries. In Singapore, construction, manufacturing and transport made up more than half of all fatal workplace accidents in 2025.

From HDB worksites to Changi’s Terminal 5

Ailytics started in 2019 as a joint research project between NUS and the Housing and Development Board (HDB). Lenard Tan (CEO) and Prateek Manocha (CTO) formally founded the company in 2021. Its early work was on HDB and JTC projects and with contractors such as Woh Hup, Obayashi and Soilbuild. It now runs large deployments at NatSteel, the PaxOcean shipyard in Indonesia, DHL and Rome’s Leonardo da Vinci International Airport. It also has more than 1,000 cameras on the Changi Airport Terminal 5 build.

Also Read: What 9 Tokyo deep tech startups learned about Southeast Asia’s appetite for Japanese innovation

“This round puts local teams on the ground in the US, Japan, Australia and the Gulf, and in the US we intend to win the mid-sized industrial companies that incumbents have priced out,” Tan said.

That is a sharp pitch. US enterprise safety software is crowded at the top end, but mid-sized manufacturers and logistics operators have often been left with clipboards and spreadsheets.

Manocha says the money will fund domain-specific models and help “keep the cost of running AI low as demand grows”. That point matters. Inference costs, meaning the cost of running AI models on live footage at scale, can wipe out the margins of vision AI companies once deployments run into thousands of feeds.

The Japan play

The most strategic name on the cap table may be Safie. It supplies 29 of Japan’s 30 largest construction companies, and it signed a memorandum of understanding with Ailytics in April 2026, witnessed by Enterprise Singapore and the Tokyo Metropolitan Government. Japan’s construction sector is dealing with an ageing, shrinking workforce, so tools that let fewer supervisors oversee more sites have an obvious appeal there. Safie’s distribution reach could do more for Ailytics in Japan than the cash itself.

A crowded field of watchful eyes

Ailytics has plenty of competition. Hong Kong-based viAct has been extending vision AI into remote pipeline corridors. Closer to home, Lytehouse has partnered with Google Cloud on AI video intelligence for workplace safety. Globally, US-based Voxel, New York-headquartered Intenseye and Dublin’s Protex AI all sell similar camera-based safety analytics. Camera makers and video management platforms are also adding analytics of their own.

Also Read: Graas raises US$17M, acquires Trustana to build smarter retail AI agents

Ailytics’s answer is that its moat lies in messy, real-world data. “The access to non-public data also enables the models to be sharper for these specific purposes, and all of this creates a truly defensible moat,” said Jeremy Tan, Co-Founder and Managing Partner at Tin Men Capital. He added that Ailytics has reached 15 markets with “very robust unit economics and client churn below 1 per cent”. Neither figure was broken down further.

The deal also fits Tin Men’s thesis of backing software that fixes operational pain in unglamorous legacy industries. The same thesis sits behind portfolio companies such as Pints AI, which raised US$5.6 million in June, and construction communications startup OnSite.

The questions still to answer

First, scale. US$7.4 million is a healthy Series A by Southeast Asian standards, but it has to pay for local teams on four continents and expansion into manufacturing, logistics, oil and gas, maritime and mining. That is a long list for one cheque.

Second, consent. A camera that can spot a missing harness can also time a toilet break. That matters especially in the Gulf, where construction relies heavily on migrant workers with limited bargaining power. The announcement says nothing about how worker data is governed, who can see it, or whether it can be used for discipline rather than protection. As industrial AI spreads across Asia’s factories, buyers, regulators and unions are likely to ask.

Also Read: Muun AI expands industrial AI push with US$700K pre-seed funding

Still, the backing is solid. With SG Growth Capital, the investment platform that has been channelling capital into NUS-linked deep tech, on board, Ailytics has a strong case as a local research spinout that found paying customers abroad. It is also the kind of story Singapore likes to tell about its push to bring 10,000 enterprises and 100,000 workers into the AI economy.

The harder job is proving that millions of hours of footage can keep workers safe without simply keeping them watched.

The post Ailytics bags US$7.4M Series A to make old CCTV cameras earn their keep appeared first on e27.