Big Money in Construction Tech Is One Feature Away

01 /The Concept

Buildots is an AI-powered construction management platform that works across project types, from single-family homes to airports.

The first step is creating a detailed build plan broken into sequential phases and work categories. All that's needed is uploading the building drawings and the required schedule – the AI engine handles the rest: generating a full task list with labor estimates and structuring a plan that identifies which work can run in parallel and which must be sequential.

Once the plan is in place and crews start work, tracking kicks in through an unconventional method: cameras clipped to workers' hard hats, recording what each crew member does throughout the shift. At the end of each shift, foremen ensure every worker plugs their camera into a computer to upload the footage.

Buildots' AI analyzes the video, cross-references it against the build plan, and marks in the plan what was done and when – attaching relevant stills from the footage to each completed item. The result is ground-truth progress data collected not from self-reporting, but from observed actions and verified outcomes.

The platform then shows clearly which phases finished on schedule and which have hit unexpected delays.

A built-in AI assistant named Dot can answer any question about progress at any point in the project, backing every response with references to the plan and linked video footage stored on the platform.

Based on actual completion times, the platform continuously recalculates the overall schedule, projecting the current best-case finish date and flagging which phases are on track, already behind, or at risk.

All of this – including access to Dot – is available to every stakeholder: site managers, foremen, project managers, the developer's planning team, and the building owner.

Buildots was founded in 2018 but only really hit its stride in 2024. It currently serves around 50 construction companies, with plans to grow its market presence fourfold this year – whether by client count or by the number of active project sites.

The startup just raised $45 million, bringing total funding to $166 million.

02 /Why It Matters

AI-driven project monitoring is appearing across multiple verticals simultaneously.

Bond ([related review](/review/ty-rastjosh-kogda-rastut-tvoi-klienty)), currently in Y Combinator, built an AI that tracks every project in a company's portfolio by analyzing internal communications and project management system records – delivering daily executive summaries and a dashboard showing which projects are on track and which are slipping.

Mesmer ([related review](/review/v-pogone-za-ii-programmistami-chut-ne-zabyli-pro-zhivogo-tehnicheskogo-direktora)), also in the current YC batch, focuses specifically on software development. Its AI CTO assistant monitors technical project execution and goes a step further than generic project tracking: it can recommend which specific engineers to reallocate to a lagging project, based on their skill profiles.

Auctor ([related review](/review/hochesh-perestat-terjat-na-jetom-dengi)), a third YC batch member, monitors custom software development projects and adds a requirement-compliance layer. Its AI automatically extracts client requirements from emails and video calls, then monitors whether the actual work matches what was agreed – essentially a digital proxy of the client watching over the developer's shoulder.

Three startups in a single YC batch all working on AI project monitoring is a signal worth taking seriously.

For all three of those startups, the source data is already digital: emails, Slack messages, Jira entries. The AI learns to interpret signals that already exist in structured form.

Buildots' core innovation is harder: creating a digital record of what's happening on a physical construction site. Attaching cameras to hard hats and training AI to read the resulting footage solves the fundamental data problem – and once that's solved, everything else becomes a more familiar software problem.

XOi ([related review](/review/tema-v-kotoroj-mozhno-i-horosho-zarabatyvat-i-horosho-prodatsja)), which raised $230 million in a single round this February, built the same kind of physical-to-digital connector for field service. A technician points their phone at an equipment nameplate; the AI identifies the make and model, suggests likely failure causes based on reported symptoms, provides schematics, and walks the tech through a diagnostic and repair sequence.

03 /Opportunities

Digitally laggard industries make for attractive startup targets – the market is often large, current inefficiency is severe, and companies will pay meaningful money for meaningful improvement.

But getting new technology into those industries requires a connection point: a simple, cheap, scalable way to create a digital picture of what's happening in the physical world, to which software can then be applied. The simpler that connection, the better.

XOi tried augmented reality glasses first – they didn't work. The business only took off when the startup switched to the same basic concept using standard mobile phone cameras.

Buildots made a similar pragmatic compromise: its cameras don't stream in real time. They record, and footage is uploaded at the end of a shift. That decision avoids the need for robust site-wide connectivity and keeps the hardware simple and inexpensive. The product works because it's deployable, not because it's technically elegant.

The broader opportunity: build technology platforms for offline and physically grounded markets, using simple digitization mechanisms that require minimal infrastructure. The lesson from both XOi and Buildots is that the winning approach is usually the most practical one – not the most sophisticated.

Siro ([related review](/review/smotri-ka-ved-takie-prodazhi-tozhe-nuzhno-uluchshat)) raised $50 million just a couple of weeks ago on exactly this logic. On the surface it's a standard AI sales coach – but the differentiation is the connection point. Most AI coaching tools analyze phone calls, video meetings, and digital messages. Siro coaches field salespeople and home-services technicians who meet clients in person. The mechanism is simple: reps record in-person conversations on a standard smartphone and send the audio to the AI for analysis.

So the question worth sitting with: in which industries is that simple physical-to-digital connection still missing – and what's the most practical way to create it?