ONHUMAN is an AI robotics company developing autonomous general-purpose humanoid robots. Our goal is to build embodied AI systems that can perceive, reason, and act in the real world. ONHUMAN is headquartered in Madrid, Spain, and this role requires 5 days/week in-office collaboration.

Team & Role Overview

Our ONVULC team is responsible for developing the core AI systems that power humanoid autonomy. We are looking for an AI Engineer, Tooling to design, build, and advance the internal software platforms, 3D evaluation interfaces, and debugging workflows that supercharge our machine learning research and model development loops.

This role focuses on engineering high-performance tooling across the model lifecycle—spanning interactive dataset curation tools, real-time 3D telemetry visualization, distributed training diagnostics, and high-velocity model testing rigs that directly accelerate the iteration rate of robot intelligence.

Responsibilities

• Developer Tooling Architecture: Design and develop developer tools, dashboards, and CLIs (Command Line Interfaces) that streamline how researchers train, test, and profile neural networks.

• 3D Visualization & Telemetry Engines: Build interfaces to visualize complex multimodal inputs (e.g., synchronized video feeds, point clouds, proprioception, token trajectories) to help researchers isolate and diagnose model regressions.

• Evaluation & Benchmarking Platforms: Advance internal evaluation frameworks, auto-labeling systems, and hardware-in-the-loop benchmarking dashboards to give researchers reliable feedback on model updates.

• Pipeline Capabilities & Latency Profiling: Improve infrastructure performance through custom debugging modules, logging libraries, and real-time tensor profiling components to track down compute bottlenecks.

• Tooling Lifecycle & Optimization: Work across the internal tooling lifecycle, from gathering feature requests from machine learning teams to prototyping high-speed UIs and deploying robust desktop or web platforms.

• Cross-Functional Fusion: Collaborate closely with modeling, pretraining, video training, RL, and robot learning teams to integrate tailored workflow improvements directly into the active autonomy stack.

• Software Rigor: Design automated testing suites, continuous integration pipelines, and monitoring tools to ensure our researcher platform remains ultra-stable.

• Paradigm Development: Contribute to the development of next-generation developer tooling paradigms optimized for multi-modal, physical, and embodied AI infrastructure.

Requirements

• AI Tooling & Platform Experience: Experience designing, optimizing, and maintaining internal toolkits, machine learning platforms, or heavy-data developer applications.

• Modern Interface Foundations: Strong understanding of building high-performance UIs, custom dashboards, or web interfaces capable of rendering real-time streaming data smoothly (e.g., TypeScript, React, Python/Streamlit/Gradio, or C++ visualization apps).

• AI Infrastructure Literacy: Strong operational understanding of machine learning loops, deep learning lifecycle steps, model evaluation procedures, and dataset indexing patterns.

• Technical Proficiency: High proficiency in Python, modern web/desktop frameworks, and familiarization with deep learning frameworks such as PyTorch.

• Experimental Rigor: Exceptional engineering rigor, extreme attention to internal customer experience (the researchers), and a data-driven approach to removing developer friction.

• Software Engineering Skills: Solid software engineering skills with a strict dedication to writing reliable, clean, maintainable, and highly documented system code.

• Autonomy & Ambiguity: Ability to operate independently, bridge gaps between software infrastructure and machine learning research, and drive clean structure into ambiguous internal requirements.

Bonus Qualifications

• Robotics Telemetry Background: Experience building 3D data visualization frameworks or simulation overlays (e.g., Foxglove Studio, Rviz, WebGL, Three.js, or custom mechatronics logging systems).

• Frontier Infrastructure & Scaled Labs: Background working within specialized ML platform/tooling teams at hyper-growth AI research companies (e.g., OpenAI, Google DeepMind, Anthropic, Meta, or xAI).

• Data System Engineering: Familiarity with heavy data lake interaction, large-scale video storage layers, parallel file systems, or fast multimodal data-loading backends.

• Profile Optimization: Technical experience optimizing GPU/CPU memory footprints, profiling distributed training workflows, or working directly with NVIDIA profiling toolkits (NSight).

Compensation

The pay offered for this position may vary based on several individual factors, including job-related knowledge, skills, and experience. The total compensation package may also include additional components/benefits depending on the specific role. This information will be shared if an employment offer is extended.