Tolle Labs Pvt Ltd
Website:
tolleroads.com
Company:
https://www.linkedin.com/company/tolle-labs
Seniority: Entry level
Industries: Defense and Space Manufacturing
Job details:
Lead AI & Autonomous Systems Engineer
Artificial Intelligence | Computer Vision | Robotics | Autonomous Systems
Location: Bengaluru, India
Experience: 4–7+ years
Company: Tolle Labs Pvt. Ltd.
Employment: Full-time
About Tolle Labs
Tolle Labs is building the intelligence layer for the next generation of autonomous machines.
We work at the intersection of Artificial Intelligence, Computer Vision, Robotics, Autonomous Navigation, UAVs, UGVs, intelligent sensing, edge computing and Physical AI. Our objective is to build systems that can see, understand, reason, navigate and act autonomously in complex real-world environments.
Our platforms span unmanned aerial vehicles, autonomous ground robots, intelligent surveillance systems and emerging robotic platforms. We are looking for an exceptional AI engineer who wants to move beyond conventional software AI and build intelligence that operates in the physical world.
⸻
The Role
We are looking for a highly capable Lead AI & Autonomous Systems Engineer with strong experience across Computer Vision, Deep Learning, perception, sensor fusion and autonomous systems.
This is not simply a Computer Vision Engineer position.
We want someone capable of architecting the AI and perception stack of an autonomous machine — whether that machine is a drone, UGV, humanoid robot, robotic platform or future autonomous system.
You should be comfortable working with data coming from:
* RGB and low-light cameras
* Stereo/depth cameras
* Thermal cameras
* LiDAR
* Radar
* IMUs
* GPS/GNSS
* Ultrasonic and proximity sensors
* Other robotic and environmental sensors
Your challenge will be to convert these sensor streams into real-time situational awareness, navigation intelligence and autonomous decision-making.
⸻
What You Will Build
You will design and develop capabilities including:
Computer Vision & Visual Intelligence
* Real-time object detection and classification
* Multi-object detection and tracking
* Person and vehicle detection
* Target recognition and tracking
* Semantic segmentation
* Instance segmentation
* Image classification
* Pose estimation
* Activity recognition
* Feature extraction and matching
* Optical flow
* Depth estimation
* Monocular and stereo vision
* Scene understanding
* Terrain classification
* Change detection
* Anomaly detection
* Visual search and re-identification
* Low-light and degraded-visibility vision
* Thermal-image analytics
* Image enhancement and super-resolution
⸻
Autonomous Navigation & Robotics AI
Develop perception and intelligence systems enabling robots to autonomously understand and navigate their environments.
Capabilities may include:
* Autonomous waypoint navigation
* Vision-based navigation
* Visual odometry
* Visual-Inertial Odometry
* SLAM
* Visual SLAM
* LiDAR SLAM
* Localization and mapping
* GNSS-denied navigation
* Dynamic obstacle detection
* Collision avoidance
* Path planning
* Motion planning
* Trajectory prediction
* Terrain awareness
* Traversability analysis
* Landing-zone detection
* Autonomous precision landing
* Autonomous docking
* Follow-me / target-following systems
* Indoor navigation
* Outdoor navigation
* Autonomous exploration
* Multi-agent navigation
* Behaviour planning
⸻
Sensor Fusion & Multimodal Perception
Build systems capable of combining information across multiple sensors.
Experience or strong knowledge in areas such as:
* Camera + IMU fusion
* Camera + LiDAR fusion
* LiDAR + IMU fusion
* GPS + IMU fusion
* Radar-camera fusion
* Thermal + RGB fusion
* Depth-camera perception
* Multi-camera systems
* Point-cloud processing
* Sensor calibration
* State estimation
* Kalman Filters
* Extended Kalman Filters
* Particle Filters
* Probabilistic sensor fusion
The objective is to give autonomous machines a reliable understanding of their position, surroundings and operating environment even when individual sensors become unreliable.
⸻
LiDAR & 3D Perception
Experience with LiDAR and point-cloud processing will be highly valued.
Potential responsibilities include:
* 3D object detection
* Point-cloud segmentation
* Point-cloud registration
* LiDAR odometry
* 3D mapping
* Terrain reconstruction
* Obstacle detection
* Occupancy maps
* Cost maps
* 3D scene understanding
* Sensor-to-sensor calibration
Experience with libraries such as PCL, Open3D or equivalent frameworks is desirable.
⸻
AI / Deep Learning
You should have strong practical experience developing, training, optimizing and deploying AI models.
Relevant areas include:
* CNNs
* Vision Transformers
* Transformers
* Multimodal AI
* Representation learning
* Self-supervised learning
* Reinforcement Learning
* Imitation Learning
* Behaviour learning
* Foundation models
* Vision-Language Models
* Vision-Language-Action models
* Generative AI
* Few-shot / zero-shot vision systems
Experience adapting modern AI models for real-world robotics applications will be particularly valuable.
⸻
Edge AI & Real-Time Deployment
Our AI doesn’t live only in the cloud.
It needs to operate inside machines, in real time, often under severe compute and power constraints.
You should understand how to deploy and optimize AI models on platforms such as:
* NVIDIA Jetson
* GPU-based edge computers
* ARM-based systems
* Embedded Linux
* Robotic compute modules
* Edge accelerators
* NPUs
Experience with technologies such as:
* CUDA
* TensorRT
* ONNX
* OpenVINO
* Quantization
* Pruning
* Model compression
* Hardware acceleration
is highly desirable.
⸻
Robotics Ecosystem
Strong exposure to robotics software is preferred.
Experience with:
* ROS
* ROS2
* Gazebo
* Isaac Sim
* NVIDIA Isaac ROS
* PX4
* ArduPilot
* MAVLink
* MAVSDK
* Robotics middleware
* Robot coordinate systems and transformations
* Navigation stacks
would be valuable.
⸻
UAV / UGV / Robotic Applications
You may work on AI systems for:
UAVs
* Autonomous reconnaissance
* Target detection and tracking
* Precision landing
* Autonomous docking
* Obstacle avoidance
* GNSS-denied flight
* Terrain-following
* Visual navigation
* Swarm perception
* Aerial mapping
UGVs
* Autonomous navigation
* Urban navigation
* Terrain analysis
* Infrastructure inspection
* Obstacle avoidance
* Mapping
* Autonomous patrol
* Human / vehicle detection
Robotic / Humanoid Platforms
* Human detection
* Human pose estimation
* Gesture recognition
* Scene understanding
* Navigation around humans
* Object recognition
* Object manipulation perception
* Visual-language interaction
* Multimodal robotic intelligence
⸻
AI Architecture & Data
The role will also involve developing the infrastructure required to continually improve our autonomous systems.
You may work on:
* Dataset creation
* Data collection pipelines
* Automated data annotation
* Synthetic-data generation
* Simulation-generated training data
* Data augmentation
* Model training pipelines
* Experiment tracking
* Model evaluation
* MLOps
* Edge model deployment
* Continuous model improvement
* Real-world validation
* Simulation-to-real transfer
⸻
What We Are Looking For
Minimum Experience
4–5+ years of strong hands-on experience in one or more of:
* Artificial Intelligence
* Computer Vision
* Robotics
* Autonomous Systems
* Machine Learning
* Deep Learning
* Robotic Perception
We care considerably more about what you have built than your designation.
⸻
Technical Skills
Strong proficiency in:
Programming
* Python
* C++ preferably
* Linux
AI Frameworks
* PyTorch
* TensorFlow
* OpenCV
Exposure to:
* YOLO
* Detectron
* MMDetection
* Hugging Face
* OpenMMLab
* NVIDIA DeepStream
* TensorRT
is desirable.
⸻
The Engineer We Want
We are especially interested in someone who can look at a problem such as:
“Make this machine understand its environment and operate autonomously.”
…and independently break that problem into:
Sensors → Data → Perception → Localization → World Model → Planning → Decision → Action.
You should be comfortable moving between:
research papers → algorithms → code → simulation → hardware → field testing.
⸻
You Will Thrive Here If You
* Love building things that move in the physical world.
* Enjoy difficult engineering problems without obvious answers.
* Can rapidly prototype and test new AI ideas.
* Understand both AI research and practical engineering.
* Are comfortable debugging software on actual robotic hardware.
* Think across software, sensors, electronics and robotics rather than operating inside a narrow AI silo.
* Can lead younger engineers and help establish strong engineering practices.
* Are comfortable owning an entire problem rather than only one algorithm.
⸻
Bonus Skills
Exceptional candidates may additionally have experience with:
* Reinforcement Learning
* Behaviour Cloning
* Vision-Language-Action models
* Embodied AI
* Foundation models for robotics
* Multi-agent systems
* Swarm robotics
* Autonomous drone systems
* Defence robotics
* Synthetic data
* Digital twins
* Sim2Real
* Nvidia Isaac
* Robot learning
* 3D Gaussian Splatting
* Neural Radiance Fields
* Neural mapping
* Event cameras
* Radar perception
⸻
What You Will Own
The person joining us should ultimately become one of the key technical owners of Tolle Labs’ AI & Autonomy Stack:
PERCEPTION
Computer Vision · LiDAR · Radar · Sensors
↓
LOCALIZATION
VIO · SLAM · Sensor Fusion · GNSS-Denied Navigation
↓
WORLD UNDERSTANDING
Objects · Humans · Terrain · Obstacles · Threats · Environment
↓
PLANNING
Path Planning · Motion Planning · Mission Planning
↓
INTELLIGENCE
Decision Making · Behaviour · AI Models · Autonomy
↓
ACTION
UAV · UGV · Robot · Autonomous Machine
⸻
Why Tolle Labs?
Most AI engineers build models that live behind a screen.
At Tolle Labs, your models could fly a drone, navigate a robot, understand a battlefield, inspect infrastructure, avoid obstacles, identify objects, map an unknown environment or autonomously complete a physical mission.
We are building Physical AI — intelligence that leaves the computer and enters the real world.
If you want to work on some of the most difficult and exciting problems across AI, robotics and autonomous systems, we want to speak with you.
Click on Apply to know more.
