Senior Machine Learning Engineer — Robotics

Senior Machine Learning Engineer — Robotics

SKEYEBOTS | Bengaluru, India | Full-Time | Hybrid

About SKEYEBOTS

SKEYEBOTS builds AI-driven autonomous systems for unmanned platforms operating across aerial, ground, and maritime environments — where perception, control, and reasoning failures are not an option.

We are a well-funded, rapidly scaling deep-tech organisation with engineering operations across India.

The Role

We're hiring a Senior Machine Learning Engineer to build and ship the perception, control, and multimodal reasoning systems that let our robots operate reliably in the real world. You'll own models end to end — from data pipeline and training through on-robot deployment and evaluation — including our work bringing vision-language models (VLMs) and vision-language-action (VLA) approaches into the robotics stack.

As a senior member of the team, you'll mentor junior engineers and provide technical guidance to 1–2 engineers working alongside you. You'll work shoulder-to-shoulder with controls, hardware, and systems engineers.

This is a hands-on role for someone who has already put ML on real robots and knows how messy the gap between a benchmark and a deployed system is.

What You'll Own

• Design, train, and deploy ML models for robot perception (object detection, segmentation, pose estimation, depth, or SLAM) and/or learning-based control (RL, imitation learning, model-based approaches)
• Adapt and fine-tune VLMs and VLA architectures for language-conditioned perception, grounding, and manipulation — and make them run within real robot latency and compute budgets
• Own the full lifecycle: data collection and labeling strategy, training infrastructure, evaluation harnesses, and on-robot deployment under latency and compute constraints
• Build evaluation pipelines that measure real-world performance, not just offline metrics — including sim-to-real transfer and failure-case analysis
• Optimise models for embedded/edge inference (quantisation, pruning, hardware-specific runtimes)
• Collaborate with controls and hardware teams to close the loop between perception, planning, and actuation
• Debug systematically across the stack when a model works in sim but fails on hardware
• Mentor junior engineers and guide the work of 1–2 engineers — through code and design review, pairing, and unblocking — helping them grow into owning their own components

What We're Looking For

Required

• 6+ years building ML systems, with meaningful experience deploying models on physical robots or autonomous systems (not just offline/benchmark work)
• Experience mentoring or providing technical guidance to other engineers — you've been the person others come to when things break
• Strong Python and PyTorch (or equivalent); comfortable with the full training-to-deployment path
• Depth in at least one core area: computer vision / perception, reinforcement or imitation learning, SLAM/state estimation, or motion planning
• Solid grounding in the practical problems of real-world ML: data quality, distribution shift, sim-to-real gap, and evaluation under uncertainty

Preferred

• Modern C++ (14/17) for performance-critical, on-robot code — deploying models via LibTorch, TensorRT, or ONNX Runtime, and working with Eigen/OpenCV
• Working familiarity with modern foundation models — transformers, and ideally VLMs — including fine-tuning, prompting, or adapting them for a downstream task
• Experience with ROS/ROS2 and robotics middleware
• Familiarity with simulation environments (Isaac Sim, MuJoCo, Gazebo)
• Edge/embedded inference optimisation (TensorRT, ONNX, on-device runtimes)
• Hands-on experience with VLA models (e.g. RT-2, OpenVLA, π0) or language-conditioned policies on real robots
• Experience fine-tuning or serving large multimodal models under real-time constraints
• Publications or open-source contributions in robotics ML or embodied AI

What You'll Get

• Competitive compensation + ESOP allocation — structured to reflect the seniority and scope of this role
• Direct collaboration with controls, hardware, and systems teams on real deployed platforms
• The mandate to shape how ML and multimodal reasoning are built into SKEYEBOTS' autonomy stack
• Mentorship responsibility and technical influence from day one
• Your work deploys into real, operational autonomous systems — not just benchmarks

Location

Bengaluru, Karnataka — hybrid, with travel to Kochi as required.

How to Apply

Send your CV and a brief note — a case where a model worked in sim but failed on hardware, and how you closed that gap — to LinkedIn Easy Apply.

We move quickly. Strong profiles hear from us within a week.

SKEYEBOTS is an equal opportunity employer. We evaluate candidates solely on capability, experience, and drive.