Career fair: 2-minute survey

* One per person for eligible majors. Requires a valid phone number.

Industrial Large Spatial Model for Real Operations.

Building Industrial Large Spatial Models.

LiOps trains models on LiDAR and camera data to infer process-critical information directly from 2D/3D inputs.

Open Roles

We are hiring exceptional teammates across engineering, research, and go-to-market. If you don't see the right role, reach out anyway—we love meeting future LiOps builders.

Full-time onsite with the Seongsu HQ team (2 minute walk from Seongsu Station).

Backed by Top Investors & Rapid Growth

Raised 900M KRW in cumulative funding from Bon Angels Venture Partners, FuturePlay, and Krew Capital—Korea's premier early-stage investors. SNU Big-Scaleup Demo Day finalist, Manufacturing Innovation Support Program (500M KRW), and recognized by government programs for our proven technology and rapid growth trajectory.

Problem-First, Tool-Agnostic

We optimize for solving customer problems, not defending paradigms. Whether it's a traditional approach or your novel method, if it makes sense, we'll try it first.

Diverse Hardware Playground

Work hands-on with AMRs, collaborative robots, forklifts, LiDAR sensors, and 3D cameras (Helios, Photoneo, Zed2i, etc.)—a dynamic environment where you'll gain breadth across the robotics stack.

Full-timeSeoul · Onsite

3D AI Engineer

3D VisionPoint CloudRepresentation Learning

Research and build industrial Large Spatial Models powering LiOps robotics across logistics and manufacturing sites.

View details

🧠 Volumetric Point Cloud Unsupervised Learning

Research and develop unsupervised learning techniques that leverage volumetric information in point clouds.

→Most industrial point cloud AI today depends on heavily labeled supervised datasets, making models brittle to sensor or environmental changes.
→LiOps builds unsupervised models with resilient representations that handle sensor diversity, domain shifts, and unseen datasets, and deploys them across real industrial projects.

🔄 Self-supervised Learning for Manufacturing Data

Conduct self-supervised pretraining on manufacturing datasets to unlock a universal 3D foundation model for robotics.

→Pursue self-supervised pretraining that can generalize across robotics use cases such as manipulators, forklifts, and digital twins.
→Design and study domain adapters that deliver few-shot generalization from learned weights.

🤖 End-to-End 3D Deep Learning for Autonomous Robotics

Develop end-to-end models that turn point clouds into actionable embeddings and control sequences for LiOps' autonomous robotics platforms.

→Generate high-quality embeddings for planners from point cloud inputs and produce situation-aware control sequences across forklifts and mobile manipulators.
→Investigate reinforcement learning and transductive learning to cover edge cases and diverse scenarios.

Qualifications

At least one project—professional or academic—related to any of the following topics (or a closely related field):

3D backbones (e.g., Point Transformer, Sparse Convolution)
Neural Signed Distance Fields
3D point cloud semantic or instance segmentation
3D object detection in point clouds
Point cloud registration
Point cloud retrieval
LiDAR-based lifelong SLAM
Online map updating and map merging
Self-supervised or unsupervised representation learning for 3D point clouds
Domain adaptation or generalization for 3D perception

Apply now

Full-timeSeoul · Onsite with frequent travel

Field Application Engineer

Customer SuccessRobotics IntegrationField Deployment

Bridge technical excellence and customer success by deploying LiOps' 3D spatial intelligence solutions at shipyards, manufacturing plants, and logistics sites.

View details

🏗️ Customer Site Deployment & Integration

Deploy and integrate LiOps' industrial Large Spatial Model solutions at customer facilities—shipyards, automotive plants, and warehouses.

→Work directly with customers to integrate 3D segmentation, registration, and reconstruction systems into their robotics workflows
→Troubleshoot on-site technical issues and adapt solutions to meet specific customer environments
→Conduct PoCs and ensure successful production deployments from day one

🔧 Technical Support & Problem Solving

Provide hands-on technical support and solve integration challenges between LiOps products and customer requirements.

→Serve as the technical bridge between customers and the core engineering team
→Identify gaps between product capabilities and customer needs, then work with engineering to close them
→Document deployment patterns, edge cases, and field learnings to improve product robustness

📊 Customer Success & Feedback Loop

Ensure customer projects succeed and capture field insights to drive product evolution.

→Monitor deployment metrics and ensure customers achieve ROI targets (e.g., 12-month payback, cost savings)
→Collect and communicate customer feedback, feature requests, and pain points to the product team
→Build long-term relationships with key customers in shipbuilding, automotive, and logistics sectors

Qualifications

We're looking for someone who combines technical depth with customer-facing skills:

Background in Robotics, Computer Vision, Mechanical Engineering, or related field (Bachelor's or equivalent experience)
Programming proficiency in Python and/or C++ (ability to debug, adapt, and integrate code)
Hands-on experience with robotics systems (AMRs, manipulators, industrial robots) or 3D vision systems preferred
Strong problem-solving skills and comfort working in unstructured, dynamic field environments
Excellent communication skills in Korean and English (business level)
Willingness to travel frequently to customer sites (domestic and occasional international)
Junior to mid-level engineers (0-3 years) encouraged to apply—field experience provides exceptional learning opportunities

Apply now

Full-timeSeoul · Onsite

Robotics Engineer

Motion PlanningAutonomous RoboticsControl Systems

Connect LiOps' 3D spatial intelligence to real robot control, building autonomous robotics systems that operate in shipyards and manufacturing facilities.

View details

🤖 Perception-to-Control Integration

Design and implement pipelines that transform 3D perception outputs into robot control commands for shipyard and manufacturing environments.

→Convert LiOps' 3D segmentation, registration, and reconstruction results into formats usable by motion planners
→Develop point cloud-based obstacle avoidance, path planning, and gripper pose estimation algorithms
→Optimize the entire Perception→Planning→Control loop to operate within <100ms latency

🚜 Multi-Platform Autonomous Systems

Deploy autonomous navigation and manipulation capabilities across AMRs, forklifts, and mobile manipulators.

→Design and tune ROS/ROS2-based navigation stacks (SLAM, path planning, obstacle avoidance)
→Integrate with industrial robot arms (UR, Fanuc, etc.) and AGV/AMR hardware
→Implement robust motion planning for unstructured environments (shipyard block yards, production lines)

🎯 Real-World Deployment & Field Testing

Validate and refine robot systems in actual customer sites, not just in simulation.

→Lead PoC and pilot deployments at shipyards and warehouses
→Design fail-safe mechanisms to handle edge cases and safety scenarios
→Iteratively improve control policies based on field data (sim-to-real transfer, online learning)

Qualifications

We're looking for candidates who meet at least 2 of the following core requirements:

Experience developing ROS/ROS2-based robot systems (navigation, manipulation, etc.)
Experience implementing motion planning algorithms (RRT, A*, trajectory optimization, etc.)
Experience integrating and debugging real robot hardware (AMRs, robot arms, AGVs, etc.)
Preferred: Point cloud-based perception-control integration experience
Preferred: Advanced control theory and implementation (MPC, LQR, etc.)
Preferred: Reinforcement learning for robotics (sim-to-real, policy learning)
Preferred: Industrial robot programming (UR Script, KUKA KRL, Fanuc TPP, etc.)
Preferred: Embedded systems and real-time control (RTOS, EtherCAT, etc.)
Preferred: Robotics simulator experience (Gazebo, Isaac Sim, etc.)
Mindset: Pragmatic focus on industrial deployment rather than lab experiments, with collaborative approach to building E2E systems

Apply now