AI Engineer – Physical AI & Swarm UAVs

Amber Wings Co; Bengaluru; Hybrid; INR 1,200,000 - 1,800,000; Full time

Job summary

Amber Wings Co. seeks a Junior AI Engineer to bridge digital algorithms and physical reality by building Physical AI systems for autonomous multi-drone swarms. You will design, train, and deploy RL models and contribute to Sim2Real transfers that land in outdoor field tests near Bengaluru.

You will transform PyTorch/TensorFlow work into deployable C++/Python packages and collaborate with robotics teams to ship robust software for flight hardware.

Qualifications

• Engineering degree in an engineering discipline (CS, Aerospace, Mechanical, ECE, Robotics, Mechatronics, etc.).
• Strong math background: Linear Algebra, Calculus, Probability & Statistics, Optimization.
• Proficiency in Python; good data structures and algorithms.
• Interest in applying AI to physical platforms (drones/robots) and willingness to travel for field tests.

Responsibilities

• RL model training, evaluation, and optimization for single- and multi-agent swarm UAVs.
• Sim2Real transfer using robotics simulators for real UAV deployment.
• Participate in field test trials near Bengaluru to collect data and validate performance.
• Translate trained models into deployable C++/Python packages for flight hardware.

Skills

Python

C++

ROS/ROS2

Field testing

Hybrid work

Education

Engineering degree

Tools

Gazebo

ROS/ROS2

PyBullet

PX4

ArduPilot

Job description

Are you passionate about bridging the gap between digital algorithms and physical reality? We are looking for an ambitious Junior AI Engineer to join our team building cutting-edge Physical AI systems.

In this role, you will design, train, and deploy Reinforcement Learning (RL) models for autonomous multi-drone (swarm) systems. You will play a hands‑on role—from training agents in high‑fidelity simulated environments to running field trials on real hardware. If you want to push the boundaries of multi‑agent autonomy and see your code fly in the physical world, this is the place for you.

Key Responsibilities

• RL Model Training & Optimization: Develop, train, and evaluate single-agent and multi-agent reinforcement learning algorithms for swarm UAV navigation, control, and mission strategy.
• Simulation to Real (Sim2Real): Utilize robotics simulators to simulate complex environments, refine reward functions, and execute Sim2Real transfer for physical UAV deployment.
• Field Testing & Data Collection: Participate in field test trials at outdoor testing sites (located on the outskirts of Bengaluru) to validate algorithm performance, troubleshoot edge cases, and collect flight data.
• Algorithm Pipeline Integration: Collaborate with software and control systems teams to translate trained PyTorch/TensorFlow models into deployable C++/Python packages for flight hardware.

Must-Haves (Basic Qualifications)

• Education: Degree (B.E./B.Tech/M.E./M.Tech) in an Engineering discipline (Computer Science, Aerospace, Mechanical, ECE, Robotics, Mechatronics, etc.).
• Core Competencies: Strong, demonstrable capabilities in three foundational pillars:
• Mathematics: Linear Algebra, Vector Calculus, Probability & Statistics, and Optimization principles.
• Computer Science: Data structures, algorithms, modular code design, and object‑oriented programming in Python.
• Artificial Intelligence: Fundamental understanding of Machine Learning and Reinforcement Learning concepts (e.g., Markov Decision Processes, Policy Gradients, Q‑Learning).
• Demonstrated Interest in Physical AI: Strong passion for applying AI to real‑world physical platforms (drones, robotics, or autonomous vehicles). Academic projects, capstone work, or personal side projects are a big plus.
• Travel & Testing Willingness: Willingness to travel regularly to field testing grounds located at the outskirts of Bengaluru for real‑world UAV flight trials.
• Work Preference: Ability to thrive in a hybrid work environment (mix of remote/office software development and hands‑on outdoor field testing).

Good-to-Haves (Nice-to-Haves)

• Robotics Simulators: Hands‑on experience with robotics or physics simulation environments (e.g., Gazebo, AirSim, NVIDIA Isaac Gym, PyBullet, Webots).
• Swarm & Multi‑Agent AI: Exposure to Multi‑Agent Reinforcement Learning (MARL) algorithms (e.g., MAPPO, MADDPG) or decentralized consensus algorithms.
• Robotics Frameworks: Familiarity with ROS / ROS2 architecture and node communication.
• UAV Hardware & Flight Controllers: Experience with open‑source flight stacks like PX4 or ArduPilot, or basic hands‑on experience assembling/debugging UAV hardware.
• Systems Programming: Proficiency in C++ in addition to Python for real‑time edge execution.

What We Offer

• Work at the Cutting Edge: Direct exposure to bleeding‑edge Physical AI technologies and multi‑drone autonomy.
• Real‑World Impact: Watch your algorithms control hardware in real‑world field environments rather than staying confined to benchmark datasets.
• Mentorship & Growth: Collaborate closely with senior AI researchers and robotics hardware engineers in an environment built for fast learning and innovation.