From hotel front desks to construction job sites, a new wave of startups are building AI that does the unglamorous work.
Foreman: The job site in your pocket
Residential contractors have been running complex businesses on a patchwork of tools that were never built for construction. Estimates live in spreadsheets. Plans travel by email. Proposals get copy-pasted from old Word documents. Foreman replaces all of it with a single AI-powered platform that manages the full lifecycle of a construction job, from first estimate to final invoice. Contractors upload plans, photos, or a project description, and Foreman’s AI generates structured takeoffs, scoped estimates, and professional proposals in minutes. Every customer record, document, permit, photo, change order, and conversation lives in one place, organized around the job. The pitch is simple: keep contractors on the job site, not behind a desk.
Aurorin CAD: The CAD software that actually opens in under four hours
Most professional CAD software still runs on kernels built in the 1980s. The consequences are real: it is common for programs like SolidWorks and Fusion 360 to take four hours to open a complex assembly file. Aurorin CAD is a ground-up rebuild, with a custom parametric and B-Rep driven kernel designed to take full advantage of modern CPUs and GPUs. The result is a tightly integrated AI agent that lets engineers create parts in seconds that would take twenty minutes in legacy tools. Founder Michael Baron is a three-time SpaceX intern who worked on Raptor combustion simulation and Dragon guidance and navigation before starting the company. He knows what it costs when engineering tools slow engineers down.
Glue: The design canvas where AI does the building
The way software gets built is changing. Coding agents are collapsing the gap between design and engineering, and the tools have not kept up. Glue is an open-source design editor built with AI as the primary user rather than an afterthought. It gives coding agents an infinite canvas to create new interfaces, iterate on existing ones, and turn changes into production-ready code in real time. Builders no longer have to work in a static mockup tool in one window and a code editor in another. Co-founder and CTO Tejas Priyadarshi previously worked as an engineer at Microsoft and Meta. The company is two people, building fast.
Lab0: The forward-deployed engineer that scales
Enterprise software implementation has been broken for decades. Every rollout, whether it is a traditional ERP or a new AI system, requires months of manual integration, configuration, and tuning from armies of system integrators. Lab0 builds agentic systems that automate this layer, compressing months-long implementations into weeks. The founders are targeting a category of work that is enormous, expensive, and almost entirely invisible to anyone who has not lived through a failed enterprise software rollout. Anyone who has knows exactly how much is at stake.
Lance: The front desk that never goes home
Hotels still run on phone calls for everything. Guests call for towels, late checkout, maintenance, parking, and directions. Each call kicks off a manual chain of tasks: someone answers, gathers details, routes the request, tracks it down, and follows up. Lance replaces that chain with autonomous AI agents that answer calls, close sales, and run operations. The company is already live across more than 50 hotels under brands including Marriott, Hilton, and Hyatt. Its agents are built on computer use technology, meaning they can navigate existing hotel software without requiring API integrations, a meaningful advantage in an industry still running on legacy property management systems. Lance raised a $3.7 million seed round in early 2026, backed primarily by hotel owners and operators.
