Quidient is a deep tech AI company pioneering advancements in Generalized (5D) Scene Reconstruction (GSR). GSR is poised to become one of the world’s great digital product categories (think GPS, MRI, and LMM). Our flagship GSR product, Quidient Reality®, is a powerful API that enables anyone with a mobile device to virtualize, visualize, and measure anything.

Words relevant to Quidient include Generative AI, Physics-Informed AI, Large Scene Models (LSMs), Large World Models (LWMs), and API-First. Overview Quidient is seeking a Senior Offline Mapping Engineer to own and advance the accuracy and robustness of our offline mapping and 3D reconstruction systems. This role sits at the core of our Generalized Scene Reconstruction Platform, where classical SfM and multi-view stereo provide the foundation and modern deep learning methods push quality over the edge — particularly in textureless, featureless, and highly reflective environments that defeat traditional approaches.

You will enhancethe pipeline that turns real-world captures into highly accurate, production-quality 3D reconstructions. This is a hybrid position, meaning that you will need to live within easy driving distance to our Technology Center in Columbia, Maryland.

What You’Ll Do

Offline Mapping & Reconstruction Develop and advance our offline mapping and reconstruction pipeline, driving accuracy and robustness across the hardest capture scenarios — textureless walls, highly reflective surfaces, featureless geometry, and large-scale scenes. Reduce pose estimation failures and increase geometric accuracy in environments where classical methods degrade, using a combination of improved estimation and learned components. Design and integrate deep learning methods (learned feature matching, monocular depth priors, learned outlier rejection) alongside classical SfM and MVS components to close the last 10% of reconstruction quality.

Improve calibration pipelines, bundle adjustment robustness, and dense reconstruction fidelity in offline processing contexts where throughput matters but hard real-time does not. Research Integration & Evaluation Build and maintain evaluation methodology grounded in real-world captures — covering feature-rich, textureless, and reflective environments — not synthetic benchmarks alone. Stay current with the deep learning and 3D vision literature, applying good judgment about which methods are production-viable and which are benchmark artifacts.

Collaborate closely with the SLAM and real-time mapping team to share components and ensure offline improvements feed back into the broader reconstruction platform.

Qualifications

: Master’s, or PhD in Computer Science, Computer Vision, Robotics, or a related field — or equivalent demonstrated experience. This is a Senior-to-Staff level role. Significant hands-on experience building or substantially improving an offline SfM, multi-view stereo, or dense reconstruction system in production — not just research prototypes.

Strong C++ and Python. Deep working knowledge of multiview geometry, bundle adjustment, nonlinear estimation, and — critically — the practical failure modes of each. Real experience with sensor calibration on real hardware.

Hands-on ability to design, train, and integrate deep learning components (learned matching, depth estimation, feature extraction) into a classical reconstruction pipeline using PyTorch or equivalent. Willingness to work on-site in Columbia, MD, in a hybrid capacity.

: Experience in fast-paced or startup environments. Prior work on reconstruction of textureless, reflective, or geometrically challenging environments. Published or shipped work combining learned and classical methods in 3D vision pipelines.

Requirements

, which may include, but are not limited to a background check, citizenship verification, and Criminal Justice Information Services verification.