Senior Robotics Perception Engineer

TelebortiX; Dadri; On-site; INR 3,000,000 - 4,200,000; Full time

Job summary

Telebortix Innovations Pvt. Ltd. in New Delhi is seeking a senior technical lead to own the perception stack behind our GNSS-denied navigation system and related ISR, mine-detection, and swarm platforms.

You will set technical direction, mentor engineers, and ship perception systems that work in the field, not just in the lab, with experience across VIO/SLAM, multi-sensor fusion, and edge compute. Role requires 6–10 years in robotics, a relevant advanced degree (M.Tech/PhD), and willingness to

Qualifications

• M.Tech/PhD (or equivalent) in Robotics, Computer Vision, or related field.
• 6+ years building perception systems for real robots/UAVs deployed outside the lab.
• Deep expertise in VIO/SLAM (e.g., VINS, ORB-SLAM, OKVIS, MSCKF-class) and multi-sensor calibration.
• Strong modern C++ and Python; ROS 2; CUDA/TensorRT deployment experience.
• Proven record of taking perception systems from research to field-deployed product.

Responsibilities

• Lead architecture and development of visual-inertial odometry (VIO), SLAM, and multi-sensor fusion pipelines (camera, IMU, radar, LiDAR, thermal) for GNSS-denied flight.
• Own camera/IMU calibration workflows (Kalibr or equivalent), sensor synchronization (PPS/hardware triggering), and deployment on edge compute (NVIDIA Jetson Orin class).
• Develop and deploy deep-learning perception: object detection/tracking, semantic segmentation, terrain-relative navigation, target geolocation - optimized with TensorRT/ONNX for real-time edge inference.
• Drive robustness: performance across day/night, thermal, degraded-visual environments, and adversarial/contested conditions.
• Mentor junior perception and software engineers; set code quality, testing, and dataset standards.
• Interface with Army/Navy stakeholders during trials; translate operational requirements into perception specs.

Skills

VIO/SLAM

C++

Python

ROS 2

CUDA/TensorRT

Multi-sensor fusion

Edge computing

Sensor calibration

Deep learning

Education

M.Tech/PhD or equivalent

Tools

Kalibr

TensorRT

ONNX

NVIDIA Jetson

CUDA

Job description

Telebortix Innovations Pvt. Ltd. - Open Positions

Location: New Delhi (on-site) | Company: Defence-tech startup building autonomous UAV systems for the Indian Armed Forces - GNSS-denied navigation, edge autopilots, swarm coordination, tactical and heavy-lift UAVs.

Note: All roles involve work on defence programmes. Indian citizenship is required. Roles involve periodic travel to field trial locations (Army/Navy establishments across India).

Team: Autonomy - Perception & Navigation | Experience: 6-10 years | Type: Full-time, New Delhi

About the Role

A senior, technical-leadership role owning the perception stack behind SARATHI, our GNSS-denied navigation system, and perception capabilities across ISR, mine-detection, and swarm platforms. You'll set technical direction, mentor engineers, and ship perception systems that work in the field - not just in the lab.

Responsibilities

• Lead architecture and development of visual-inertial odometry (VIO), SLAM, and multi-sensor fusion pipelines (camera, IMU, radar, LiDAR, thermal) for GNSS-denied flight.
• Own camera/IMU calibration workflows (Kalibr or equivalent), sensor synchronization (PPS/hardware triggering), and deployment on edge compute (NVIDIA Jetson Orin class).
• Develop and deploy deep-learning perception: object detection/tracking, semantic segmentation, terrain-relative navigation, target geolocation - optimized with TensorRT/ONNX for real-time edge inference.
• Drive robustness: performance across day/night, thermal, degraded-visual environments, and adversarial/contested conditions.
• Mentor junior perception and software engineers; set code quality, testing, and dataset standards.
• Interface with Army/Navy stakeholders during trials; translate operational requirements into perception specs.

Required Qualifications

• M.Tech/PhD (or equivalent experience) in Robotics, Computer Vision, or related field.
• 6+ years building perception systems for real robots/UAVs deployed outside the lab.
• Deep expertise in VIO/SLAM (e.g., VINS, ORB-SLAM, OKVIS, MSCKF-class filters) and multi-sensor calibration.
• Strong modern C++ and Python; ROS 2; CUDA/TensorRT deployment experience.
• Proven record of taking perception systems from research to field-deployed product.

Preferred

• Experience with radar or thermal-based perception.
• Track record of leading small technical teams.