Origin, a full-stack physical AI company building general-purpose robots for interior construction, emerged from stealth on Thursday with $20 million in total funding, co-led by Accel and Zacua Ventures. Other backers include Unilazer Ventures, Schema Ventures and Z21, alongside angel investors from Anduril, Skild AI, 1X and Tesla.
The funding will primarily go towards deploying OG-1, Origin's flagship robot, and building out its fleet. It will also support large-scale pilots with customers in the US and Europe. These include two of the largest drywall-finishing contractors in the US, with additional pilots scheduled in Germany, Sweden, France and Saudi Arabia, and will extend the platform to new interior trades.
“Too many construction robots perform well in a lab and fail within a week on a live job site. Origin built for the job site from the start, and it shows in the results: the robot works, crews trust it and the company can manufacture it at volume,” said Vivin Hegde, founding partner at Zacua Ventures.
OG-1 is Origin’s flagship robot, designed to perform multiple interior construction trades on a single platform. Currently, it autonomously completes wall-finishing workflows without human intervention. It applies compound, sands and primes walls to a professional finish.
Also Read
OG-1 moves on a three-wheeled platform and has a 360-degree view of its surroundings. It can work at heights of up to 14 feet and handle contact-rich tasks, such as sanding, using real-time force control to maintain consistent pressure against the wall. OG-1 is designed to work safely alongside construction crews while other trades continue on site, without requiring a separate cordoned-off work area.
“We believe a foundational physical AI model built specifically for construction is the right approach to win in this massive market, where robotics has barely scratched the surface,” said Anagh Prasad, investor at Accel.
OG-1 is powered by Origin’s Construction Action Model, a foundation model built specifically for construction. It brings together three core components. These include a reasoning system that plans and orchestrates complex subtasks in construction workflows. There is an action policy that enables OG-1 to navigate autonomously, identify exactly where to work, determine how to use each tool and assess the quality of its work. There is also a controller that translates the policy’s actions through precise, feedback-driven control of the robot.
The model is trained using high-fidelity simulation and deployment data from live job sites.
Origin has been deployment-focused from the beginning and deployed its first robots on job sites in New York within 10 months of its founding. Currently, OG-1 is being deployed across commercial spaces, multifamily apartments, data centres, hotels and warehouses.
“Over the last century, construction has only mechanized or automated heavy exterior work. But much of the skilled work inside the building, which accounts for more than 65 per cent of the construction timelines, is still done by hand. We started Origin to bring physical AI to interior trades,” said Yogesh Ghaturle, co-founder and CEO of Origin.
Zacua's Construction Robotics Report 2026 found that the sector attracted $1.36 billion in venture funding in the first three quarters of 2025, up 125 per cent from the $612 million raised in all of 2024, while robots still account for less than 0.03 per cent of global construction spending.
Origin's platform is ready for manufacturing at scale, and the company is targeting the deployment of 150 robots by the end of 2027. It plans to support additional interior trades such as taping, drywall installation, drilling and conduit installation, using the same hardware platform and Construction Action Model.
Origin was founded in early 2025 by Yogesh Ghaturle, Ashish Daga and Tushar Makkar. Yogesh previously founded Perpule, which Amazon acquired in 2021. Makkar heads autonomy and brings a decade of experience building AI systems at scale, most recently leading AI teams at Eightfold AI. Daga leads hardware and previously built precision automation hardware at Morphle Labs.
“We spent 18 months building before coming out of stealth because we wanted to prove what our robots could do on real job sites,” added Ghaturle. “Today, our robots are already working at three times human speed, delivering human-level quality and generalising across different site conditions.”