Why One Safety Founder Calls WhatsApp the Most Successful Construction Tech Ever Built

Walk onto most sites and the safety tool people actually use is WhatsApp. The free chat app in their pocket, not the platform IT paid for. Group chats where a foreman fires off a photo of a missing guardrail and gets an answer before the morning briefing. That’s the uncomfortable starting point for Prakash Senghani’s conversation on the Bricks & Bytes podcast. He’s co-founder and CEO of Navatech, and he’s deployed enough software on major projects across the UK, Asia, and the Middle East to know what gets used and what gets ignored.

His reason for building in safety is personal. His dad was a carpenter who had to stop working in his mid-40s after back injuries. Years later, rolling out shiny tools on big jobs, Senghani kept asking whether any of it would have kept his dad working. The answer was no. Safety kept turning up as an afterthought, a form buried inside someone else’s project tool.

If you run projects, build ConTech, or just want to know why construction adopts software so slowly, this one goes straight to the part most vendors skip: the bit where you have to get real people to use the thing.

The costs that never hit your balance sheet

Productivity, morale, and people walking off the industry

Senghani’s first point is one finance teams hate, because you can’t put a clean number on it. The real damage from weak safety, he says, hides: lost productivity, sinking morale, and workers leaving an industry that feels dangerous. With a global labour shortage already biting, the perception that construction isn’t safe quietly feeds the recruitment problem.

Then there’s the measurement trap. Safety spend is hard to justify because you can’t prove a negative. When nothing goes wrong, it’s nearly impossible to say which control stopped the accident. That’s not unique to software, it haunts the whole safety profession, and it makes ROI conversations genuinely awkward. It’s also why investors keep backing the firms that can show real adoption, like safety platforms raising serious money on the promise of safer sites.

Compliance versus capacity

Looking backward, and building the muscle to manage risk

Senghani splits safety into two jobs. The first is compliance, and it mostly looks backward. Someone cuts their finger, you write an incident report, you run a root-cause analysis, you log daily observations of near-misses so you can spot patterns before they turn into something worse. Historically that’s a paper-heavy slog, or a digital version of the same paper.

The second he calls capacity to manage risk, the leading indicators that lift how well an organization handles danger before it shows up. Training is the obvious one. The catch is how it’s usually done: herd everyone into a room, run an English-language slide deck, hand out a quiz people copy off their mates, collect signatures. A tick-box exercise dressed up as prevention.

Technology changes that by tailoring the medium and the language. Some workers want to read it, some want a video, some want it in their own language on the bus to work. Get that right and people actually absorb it, which is the whole point.

Watch the full conversation

Prakash Senghani on Bricks & Bytes

This episode is part of the Bricks & Bytes state of construction safety technology series, digging into adoption, AI on site, and what a solved safety problem looks like.

Digitalization is not transformation

When a digital form is slower than a pen

About a decade ago the industry “digitalized,” which mostly meant rebuilding paper forms as digital forms. Senghani is honest about the downside. Sometimes the digital version took longer to fill in than scribbling on a page: log in, find the form, work the dropdowns. Worse, easy data entry flooded systems with low-quality records, and the signal drowned in the noise.

What’s shifting now is the input method itself. Voice lets people log things hands-free. Vision feeds in from CCTV and 360 cameras that get analyzed automatically. Smart glasses are early but promising, and Navatech has tested Meta glasses that drop messages straight into a WhatsApp group. Senghani is even eyeing four-legged camera robots that patrol sites at any hour, the same shift toward machines that sense rather than manipulate playing out across construction robotics in 2026.

The common thread is AI closing the gap between where data is captured and where the decision gets made.

Meet workers where they already are

Free, cross-company, and zero training

Here’s the WhatsApp logic. Two companies on one project usually can’t share a licensed platform, because each bought its own software under its own agreement. But they can both jump into one group and act like the team they are on that job. It’s free, it’s simple, and nobody ever sat through training to use it.

So rather than drag people off it, Navatech put its AI inside the chat. Workers raise observations and incident records, ask questions of their company’s safety documents, or send a photo and ask what risks it shows, all in their chosen language, by voice if they prefer, without leaving WhatsApp. When something needs more structure, the system pulls them into Navatech’s own interface. Pricing is per project, not per user, on purpose, so nobody gets left off the platform because their employer wouldn’t pay for a seat.

One detail stuck with him. A customer told him the site WhatsApp chat beats his CCTV feed, because the camera can’t see around corners and a person can, and the camera can’t smell gas in a basement.

What actually protects a ConTech company

Not the code, the problem and the distribution

For the founders listening, Senghani had a clear answer on defensibility. He’s not losing sleep over a frontier lab like Anthropic shipping a construction safety product, because the big labs build platforms, not niche vertical apps. The threat he watches more closely is his own customers using Claude to build their own internal tools.

That rarely sinks a business, he says, but it does stretch the sales cycle, because now there’s a quick in-house prototype to argue against. His view on moats is refreshingly old-fashioned: the technology was always replicable with enough time and money, even before AI. What’s hard to copy is a deep understanding of the problem plus real distribution, and the unglamorous work of security, access control, and maintenance.

He’s seen it play out. A major contractor builds its own safety tool, the person who built it leaves, and the code quietly rots because nobody can update it. Vibe-coding a prototype is easy. Running it in production, for years, is not.

How site safety actually gets captured

A quick comparison of the approaches Senghani describes

Related Articles
HammerTech Secures $70 Million Investment to Drive International Expansion
Construction Robotics in 2026: 4 Workflows Where Robots Are Actually Working
What Is Construction Technology (ConTech)?