How Ailytics Took Singapore’s Vision AI to the World

Every year, nearly three million workers die from work-related accidents and diseases. That is the stark figure from the International Labour Organization. So how do you stop the next accident before it happens? One Singapore startup thinks the answer is already hanging on the wall. Ailytics takes the CCTV cameras that sites already own and turns them into smart safety watchers.

In fact, it has grown from 70 projects in four countries to more than 400 deployments across 15 countries.

In this story, you will see how Ailytics started, how its tech works, and how it raised money. Then you will learn how it went global and what you can copy for your startup. Let’s dive in.

The Ailytics Origin Story: From a Campus Project to a Startup

Most startups begin with an idea. Ailytics, however, was founded to address a real problem encountered on actual building sites. Back in 2019, the founders worked on a research project at the National University of Singapore (NUS).

The project teamed up with the Housing and Development Board (HDB), the agency that builds most of Singapore’s public homes. AI Singapore also took part.

Meet the Founders

The company formally launched in 2021. Lenard Tan (full name Tan Wei Zhuang) became CEO. Before that, he built planning software at Boeing and Jeppesen. He also earned an MBA from NUS in 2021. Meanwhile, Prateek Manocha took the CTO role and led the tech side.

So why safety? Tan has said that manual safety checks are slow, costly, and often miss things. Think about it. A safety officer can only stand in one place at a time. Yet a big site may have hundreds of workers, cranes, and trucks moving at once.

As a result, many hazards go unseen until something goes wrong.

Why Construction Came First

Ailytics started where the pain was sharpest. Construction is one of the most dangerous fields in the world. Furthermore, HDB sites gave the team a live test bed right at home. Early on, the startup built more than ten ready-made AI modules.

According to IMDA, these modules cover about 90% of common site scenarios. Plus, Ailytics joined IMDA’s Accreditation program, which helps local tech firms win trust from big buyers.

Unique insight: Here is what most coverage misses. Starting with a government partner did more than provide Ailytics with data. It also gave the team a reference client that private builders already trusted.

In other words, the HDB link worked like a seal of approval. For a young Singapore deep tech startup, that kind of proof is often worth more than the first cheque.

How Ailytics Vision Works: AI Turns Old CCTV Into a Safety Coach

Here is the clever part. Ailytics does not ask clients to buy new cameras. Instead, its software plugs into the CCTV systems a site already has. It then keeps an eye on every feed, both day and night.

The Ailyssa Platform

The main product is called Ailyssa. It is a computer vision CCTV analytics platform built for heavy industry. For example, it can spot a worker standing under a lifted load.

It can also flag a missing barrier, a person without a helmet, or a truck in the wrong lane. When it sees a risk, it sends an alert straight to the safety team’s phone or chat app.

Even better, Ailyssa can work out 3D space from a single 2D camera. So it can judge distance and depth without extra sensors. That matters because extra hardware means extra cost and a longer setup.

The system also saves short clips from before and after each alert. As a result, safety officers can use real footage in their daily briefings. Workers then see their site, not a generic training video.

AilyssaX and Smarter Reasoning

In June 2026, Ailytics rolled out AilyssaX. This is a reasoning engine, not just a detector. In simple terms, it can check whether a multi-step task follows the right order. It also keeps working in rain, dust, and low light.

Those are the harsh outdoor conditions that trip up many vision systems.

What This Means for Buyers

Early results looked strong. Ailytics said its tools helped clients find seven times more hazards. Meanwhile, manual inspections dropped by up to half.

Unique insight: Think of Ailytics as a “brownfield AI” company. Most sites already paid for cameras that nobody watches. So Ailytics is not selling a new tool. Instead, it sells value from money the client has already spent.

That makes the “yes” much easier for a careful site manager. If you are researching AI video analytics for construction, this “no new hardware” angle is the real reason the model scales.

The Ailytics Funding Journey and the Numbers Behind It

Good tech needs fuel. So let’s look at how Ailytics paid for its growth.

The Pre-Series A Round (2024)

In April 2024, Ailytics raised US$2.7 million in an oversubscribed pre-Series A round. Tin Men Capital, a Singapore venture firm, led the deal. At that point, the startup had about 70 projects in four countries.

Before that, its backers had included Startup Wise Guys, Plug and Play Malaysia, AngelCentral, and the CEMEX Ventures startup contest.

The Series A Round (2026)

Then, on 26 October 2026, Ailytics announced a US$7.4 million Series A. Once again, Tin Men Capital led, and once again, investors oversubscribed the round. This time, SG Growth Capital joined in.

So did Safie, Japan’s largest cloud camera firm, plus one of Japan’s biggest construction groups. Finally, family offices and returning investors rounded it out.

The Ailytics Growth Numbers

So what did investors see? Here are the key figures the company shared:

• 400+ deployments across 15 countries
• 5.7 million+ hazards detected
• 500 million+ square meters of worksites monitored
• 73% drop in total recordable incident rates, as reported by clients
• 84% fewer safety non-compliances
• Under 1% client churn, according to Tin Men Capital

The last number warrants further examination. Churn means the share of clients who leave. Below 1% is rare for any software company. In practice, it tells you that once a site installs Ailytics, it rarely turns it off.

Unique insight: Notice the gap between the two rounds. Ailytics grew its project count about six times in roughly two and a half years. Yet the Series A was still modest by Silicon Valley standards. This points to a capital-light model.

Because the software rides on existing cameras, each new site costs little to add. For anyone tracking industrial AI startup funding, that efficiency is the headline, not the cheque size.

How Ailytics Took Its Vision AI to the World

Now for the big question. How does a startup from a small city-state win clients on four continents? Ailytics used a mix of partners, flagship wins, and focused markets.

Ailytics Big-Name Reference Projects

First, the company built a strong client list. In Singapore, it worked with builders such as Woh Hup, Obayashi, and Soilbuild. It also runs more than 1,000 cameras on the Changi Airport Terminal 5 build.

Beyond Singapore, its clients include NatSteel, the PaxOcean shipyard in Indonesia, DHL, and Rome’s Leonardo da Vinci Airport. Each win made the next sale easier.

The Japan Play

Next came Japan. In April 2026, Ailytics signed a deal with Safie, with Enterprise Singapore and the Tokyo Metropolitan Government as witnesses. Here is why it matters.

Safie supplies 29 of Japan’s 30 largest construction firms. So instead of knocking on each door, Ailytics gained a route to almost the whole top tier at once.

The US and the Gulf

Meanwhile, the Series A funds local teams in the US, Japan, Australia, and the Gulf. In the US, Ailytics is aiming at mid-sized industrial firms. Tan says big incumbents have priced these firms out.

In the Gulf, it already has projects in the UAE and Saudi Arabia, where huge building programs are under way. As Tan put it, “We are building a global industrial AI company from Singapore.”

New Industries

Additionally, Ailytics is expanding beyond the construction industry. It now plans to grow in manufacturing, logistics, oil and gas, maritime, and mining. These fields share the same mix of heavy machines, outdoor sites, and high risk.

Unique insight: Most articles frame the expansion as a simple “Singapore to the world” leap. But look closer, and you see a pattern. Ailytics enters each market through a trusted local partner. In Singapore, that was HDB. In Japan, it is Safie.

In the US, it is the overlooked mid-market. This approach is the real global expansion strategy for startups hidden in the story: borrow trust first, then sell.

What Founders Can Learn From Ailytics

The Ailytics story is not just news. It is a playbook. Here are five steps you can apply to your startup.

Step 1: Start With a Painful, Costly Problem

Ailytics did not chase a trend. Instead, it chose workplace deaths and injuries, a problem with real human and monetary costs. So ask yourself: what problem keeps your buyer up at night?

Step 2: Use What Your Customer Already Owns

Clients did not need new cameras. That cut the cost and the risk of saying yes. Therefore, look for ways your product can ride on tools, data, or habits your buyers already have.

Step 3: Win One Trusted Reference Client

The HDB link opened doors. Likewise, a single respected client can do more than a big ad budget. So aim to land a name that others in your field look up to.

Step 4: Expand Through Partners, Not Just Sales Teams

The Safie deal shows the power of the right partner. Instead of building a huge sales force, find a partner who already reaches your buyers.

Step 5: Keep Your Costs Under Control

Running AI on thousands of live video feeds is expensive. In fact, these running costs can eat a vision AI firm’s margins. That is why CTO Prateek Manocha wants in-house models that keep costs low as demand grows. So plan your unit economics early, before scale makes them painful.

The Challenges Ahead

Of course, the road is not empty. Ailytics faces rivals such as viAct in Hong Kong, Invigilo in Singapore, and Protex AI. Plus, US and Gulf buyers will expect local support and strict data rules. Still, its low churn and real results provide it a strong base.

Unique insight: The best lesson may be patience. Ailytics spent about two years in research before it became a company. Then it spent years proving results at home. By the time it went global, it had data, not just a pitch.

For anyone building Vision AI for workplace safety, that patience is the real moat.

Conclusion: Why the Ailytics Story Matters

The Ailytics journey shows how a small team can take on a huge global problem. It began as a university project with HDB in 2019. Then it became a company in 2021. Since then, it has turned ordinary CCTV into a smart safety layer for heavy industry.

Along the way, Ailytics made a few smart moves. First, it used cameras that clients already owned, so buying was easy. Next, it proved results at home on public housing sites and the Changi Terminal 5 build.

Thereafter, it raised US$2.7 million in 2024 and US$7.4 million in 2026. Finally, it used partners like Safie to open new markets fast.

The results speak loudly. Today, Ailytics runs more than 400 deployments in 15 countries. Clients report far fewer incidents and safety lapses. And almost none of them leave.

So what is the main message? Great startups do not always need new hardware or giant budgets. Instead, they need a real problem, a trusted first client, and a smart path to scale. Ailytics is proof that a Singapore deep tech startup can compete on the world stage.

Your next step: If you run a worksite, check what your cameras could already tell you. And if you are a founder, pick one lesson from this story and test it.

Frequently Asked Questions

1. What does Ailytics do, and how does its vision AI technology work?

Ailytics builds vision AI software that plugs into existing CCTV cameras. It watches live feeds, spots hazards like workers under lifted loads, and then sends real-time alerts to safety teams. No new cameras are needed.

2. Who founded Ailytics, and when did the Singapore startup first begin?

Lenard Tan and Prateek Manocha founded Ailytics in 2021. However, the idea began in 2019 as a research project at the National University of Singapore, working with the Housing and Development Board.

3. How much total funding has Ailytics raised from its investors so far?

Ailytics has raised about US$10.1 million across its two main rounds. That includes a US$2.7 million pre-Series A in 2024 and a US$7.4 million Series A in 2026, both led by Tin Men Capital.

4. Which countries and industries does Ailytics Vision AI now serve?

Ailytics runs more than 400 deployments in 15 countries, including Japan, the US, Australia, the UAE, and Saudi Arabia. It serves construction today and is now growing into manufacturing, logistics, maritime, and mining.

5. How is Ailytics different from other AI video analytics companies?

Unlike many rivals, Ailytics reads 3D depth from a single 2D camera. Its AilyssaX engine also checks multi-step tasks in rain, dust, and low light. Plus, clients report a 73% drop in recordable incidents.