About BuildotsBuildots is the world’s most advanced construction intelligence platform. By comparing schedules and BIM against site reality, its advanced AI transforms granular data into insights that unlock operational excellence.Fortune 500 contractors and owners trust Buildots, including HOCHTIEF, Bouygues, Intel, NCC and VINCI Construction.Its impact is proven at every scale, from multi-family projects to multi-gigawatt data centers.About the RoleOur algorithms team is responsible for all of the deep learning and computer vision challenges the company is facing, tackling complex real-world problems that directly affect the company’s success. You’d be joining to lead a team of experienced researchers leading the research and implementation of our algorithmic engine analyzing large streams of visual data from construction sites, together with 3D models - This is still mostly a hands-on role.What you'll do:Take ownership on a major part of Buildots analysis pipeline - analyze week spots and build roadmap for a small team of senior researchesResearch and design deep learning solutions for different computer vision problems, using tremendous amounts of visual and 3D data.Design, research and evaluation of our production photogrammetry pipeline — feature matching, SFM and multi-view stereoFace challenges in areas of 3D Reconstruction and localization for indoor environmentsOwn the complete research cycle, from requirements to design, implementation, and benchmarking, producing innovative solutions for core company problemsRequirements:1+ years of experience mentoring or leading a small team of DL/CV researchers.7 years of experience researching deep learning for computer-vision problems.Deep expertise in 3D perception, specifically within feature matching, SfM, SLAM, MVS, depth estimation, or 3D reconstruction.MSc. in CS/EE/Similar departments.Experience with deep learning libraries (preferably PyTorch)Proficiency in Python.Ability to thrive in a fast-paced startup working environment.Good vibes and a sense of humor - must.Experience with 3D foundational models (VGGT/Depth Anything) or semi-dense/dense matchers (RoMA/DKM) - Advantage*To support our hiring process, we may leverage AI-powered tools for tasks such as application screening, resume analysis, and response evaluation. All decisions remain under the oversight of our recruitment team.*By submitting your application, you agree that Buildots will process your personal data in accordance with Buildots' Privacy Policy.