---
title: 'Deep Learning Engineer - World Models at Humanoid'
canonical: 'https://feeny.ai/job/deep-learning-engineer-world-models-humanoid-london-xqn52bhtetbc'
type: 'job'
last_seen: '2026-09-10'
---

# Deep Learning Engineer - World Models at Humanoid

- **Company:** [Humanoid](https://feeny.ai/companies/humanoid)
- **Location:** London, United Kingdom
- **Employment:** full-time
- **Work type:** onsite
- **Posted:** 2026-07-16
- **Last confirmed live:** 2026-09-10
- **Apply:** https://jobs.ashbyhq.com/humanoid/be9a09e4-768e-474a-a0d9-391b327b0ccf/application

**Skills:** Python, PyTorch, JAX, Deep Learning, LLMs, VLMs, Image Generative Models, Video Generative Models, Distributed Training, Checkpointing, State Management, Software Engineering Practices, Robotics, Autonomous Driving, RL to LLMs, VLA (Vision-Language-Action) Models, Productization of Deep Nets, OpenVLA, Physical Intelligence (π) Models

> The VLA Pre-training Engineer will develop and train capable policies for robot vision-language-action systems. Responsibilities include post-training via behavior cloning and reinforcement learning, curating data collections, and building pipelines for synthetic data and teleop logs.

## Job description

Here at Humanoid, we believe in a future where robots amplify human potential. That’s why we’ve set out on a mission to build the world’s most capable, commercially-scalable, and safe humanoid robots. We’re bringing that mission to life with HMND‑01 Alpha - our rapidly developed humanoid platform now running in real industrial pilots - and we’re growing the team to take it even further.

## Our Mission

At Humanoid we strive to create the world's leading, commercially scalable, safe, and advanced humanoid robots that seamlessly integrate into daily life and amplify human capacity.

## About the Role

As a Research Engineer on the World Models team, you will build action-conditioned generative models that predict how the world evolves around our robots — future video, proprioception, contacts, and outcomes — from past observations and actions. World models serve four purposes in our stack: a pretrained, physics-aware prior for our VLA policies; an engine for rare data collection, cross-platform transfer, and sim-to-real transfer; a testbed for policy evaluation and testing before hardware; and a future-prediction rollout engine that surfaces what our policies intend to do, for safety and planning. This is a hands-on individual contributor role: you will design architectures, run large training jobs, and validate your models against real fleet data from industrial deployments.

## What You'll Do

- Design and train multimodal world models — video, state, action, and language — using diffusion-based and transformer architectures.
- Build action-conditioned video prediction and dynamics models that stay physically consistent over long horizons, including contact-rich manipulation, and serve as pretrained priors for VLA policies.
- Develop learned-simulator evaluation: score candidate policies offline, predict real-world success rates before deployment, and roll out policy futures to expose intended behaviour for safety review and planning.
- Generate synthetic rollouts and counterfactual experience — including rare events, cross-platform transfer, and sim-to-real transfer — to augment policy training, and measure their effect on downstream task performance.
- Establish fidelity metrics and calibration protocols that quantify where the world model can be trusted and where it diverges from reality.
- Build data pipelines that turn fleet telemetry, teleoperation logs, and internet-scale video into training corpora for world models.
- Run scaling and ablation studies on architecture, data mixture, and context length; communicate findings crisply.
- Collaborate with pretraining, RL, and manipulation teams to integrate world models into policy training and evaluation loops.

## What We're Looking For

- A track record of training large generative models — video, world, or multimodal — with shipped models or published artifacts to show for it.
- Deep hands-on experience with modern generative architectures: diffusion models, autoregressive transformers, latent-variable models, or video prediction.
- Experience with large-scale distributed training: streaming datasets, checkpointing and state management, debugging numerics and training instabilities.
- Strong Python + PyTorch/JAX; you can profile kernels, optimize data loaders, and write maintainable research code.
- Empirical rigor: you design careful evaluations, run honest baselines, and document experiments clearly.
- Excitement about grounding generative models in physical reality rather than pixels alone.

## Nice to have

- Experience with world models for robotics or autonomous driving (e.g., action-conditioned video models, learned simulators, model-based RL).
- Familiarity with robotics simulators (Isaac Sim, MuJoCo) and sim-to-real considerations.
- Experience using world models for policy evaluation or synthetic data generation at scale.
- Publications at top-tier deep learning conferences (NeurIPS, ICML, ICLR, CoRL, CVPR) or equivalent open-source contributions.
- Experience optimizing generative models for fast inference.

## What We Offer

- Competitive equity: stock options with meaningful upside as we scale.
- 30+ paid days off, including 23 days of annual leave, all UK bank holidays, and additional company closure days (including Christmas–New Year shutdown).
- Private healthcare, including virtual and in-person care.
- Pension scheme with 8% total contribution (5% employee, 3% employer) on full earnings.
- Free daily breakfast, catered lunch, and snacks in-office.
- Work at the frontier - collaborate daily with world-class engineers, researchers, and product experts building the next generation of AI and humanoid robotics.
- Real ownership - direct access to founding leadership, meaningful input on product direction, and the ability to drive key initiatives from day one.

## About Humanoid

## Company Overview
- **One-liner**: Humanoid builds commercially scalable, safe, and reliable humanoid robots designed to automate labor-intensive industrial tasks across warehousing, logistics, manufacturing, and retail.
- **Entity Type**: Private (Seed stage)
- **Headquarters**: London, United Kingdom (HQ) with offices in Cambridge, MA, USA; Burnaby, BC, Canada; and additional presence in Germany, Georgia, and the UAE.
- **Founded**: 2024
- **Founders**: Artem Sokolov (CEO & Founder)

## Core Business
- **Primary industry/industries**: Robotics Engineering, Humanoid Robotics, Industrial Automation, Artificial Intelligence
- **Target customers**: B2B / Enterprise — industrial sectors including retail, e-commerce, third-party logistics (3PL), manufacturing, and automotive.
- **Mission or purpose statement**: “Empowering humanity by building the most reliable, safe, and helpful humanoid robots.” The company aims to free people from repetitive, dangerous, and tedious work, allowing them to engage in more creative and meaningful activities.

## Products & Services
- **HMND 01 Alpha Wheeled**: A robust industrial-grade wheeled humanoid robot designed for material handling, pick-and-place, and visual inspection in warehouses and factories. Prioritizes reliability, low total cost of ownership (TCO), and minimal maintenance.
- **HMND 01 Alpha Bipedal**: A bipedal humanoid robot for more complex environments requiring advanced locomotion, navigation of confined spaces, and higher dexterity. Features human-level manipulation speeds and walking speeds up to 1.5 m/s.
- **KinetIQ**: An AI framework for orchestrating fleets of humanoid robots across both wheeled and bipedal platforms, enabling centralized control and task optimization.
- **Modular Garments**: Exchangeable protective “garments” that allow HMND 01 robots to adapt to different use cases and environments, minimizing contamination and collision damage while enabling functional customization.

## Market Standing
- **Valuation/Market Cap**: Not disclosed.
- **Key Metric**: Total funding of **$50 million** (Seed round closed October 2025, with 1 lead investor).
- **Notable Investors/Partners**: Strategic partners include leading global companies in AI, cloud computing, and research. Specific partner names are not publicly listed on the website, but the company highlights collaboration with commercial partners to scale deployment.
- **Growth Signals**:
  - Headcount grew **139.2% YoY** to ~182 employees (as of mid-2026), with 65 active job postings.
  - Opened offices in London, Boston, and Vancouver within its first year.
  - Team includes talent recruited from major robotics and AI firms: Sanctuary AI, Boston Dynamics, Dyson, Arrival, Ocado Technology, 1X, Wayve, Apptronik, and Brain Corp.
  - Focus on initial target sectors representing a combined $12 trillion in annual payroll and 250 million workers.

## Competitive Advantages
- **Human-centric design**: The world is already designed for humans, so humanoids can seamlessly integrate into existing environments without costly redesign.
- **Modularity**: HMND 01’s hardware/software modularity enables multi-task performance and easy repurposing, reducing the need for specialized robots.
- **Reliability & low TCO**: Prioritizing hardware and software stability over raw performance to deliver industrial-grade robustness suitable for mass commercialization.
- **KinetIQ fleet orchestration AI**: Proprietary framework for managing mixed fleets of wheeled and bipedal robots, differentiating from competitors focused on single-form-factor robots.
- **Experienced leadership**: CEO Artem Sokolov previously scaled a family manufacturing business into a billion-dollar enterprise (EY Entrepreneur of the Year 2021). The CTO, Jarad Cannon, has extensive experience translating research into deployment-ready robotics systems.

## Strategic Focus
- **Near-term (2027)**: Deploy robots for physical tasks in manufacturing, warehousing, logistics, and retail — tackling repetitive, dangerous, and tedious manual work.
- **Medium-term (2029)**: Expand into the service sector, targeting the 70-80% of the global economy that will be services, with a focus on assisting an aging population (1.4B people over 60).
- **Long-term (2031+)**: Enter households as assistants and companions, potentially revolutionizing everyday life across 3.5 billion homes.
- **Core strategy**: Partner with leading global companies in AI, cloud, and research for technology development, and with commercial partners for scaled deployment. The company believes the humanoid TAM will reach $38 billion by 2035 and $1 trillion by 2050.

## Why Work Here
- **High-growth environment**: Company grew headcount by 139% in a single year, indicating rapid scaling and opportunity for career advancement. 65+ open roles suggest strong hiring momentum.
- **Talent from top robotics firms**: Team includes alumni from Boston Dynamics, Sanctuary AI, Dyson, Ocado Technology, Arrival, Wayve, and others — offering a deep learning and collaboration environment.
- **Technical focus**: Engineering and technical roles represent 38% of the workforce, with senior-level hires at 23% of the team. The company emphasizes hardware, AI, machine learning, and control systems.
- **Mission-driven**: Clear purpose to automate undesirable, unsafe, and repetitive work — appealing to candidates who want to make a tangible societal impact.
- **Global presence**: Offices in London (HQ), Cambridge (USA), and Burnaby (Canada), plus operations in 16 countries — offering potential for international mobility and remote collaboration.
- **Innovation culture**: Described as a place where engineers and researchers “move from demo to deployment” and “not just in the lab.” The CTO emphasizes building systems that perform in the real world from day one.
- **Office/Hybrid policy**: Not explicitly stated, but the company has physical offices in three countries and a distributed team across 16 nations — likely supports hybrid/remote with in-office collaboration for engineering roles.

## Sources
1. [thehumanoid.ai](https://thehumanoid.ai/)
2. [thehumanoid.ai/about-us/](https://thehumanoid.ai/about-us/)
3. [thehumanoid.ai/team/](https://thehumanoid.ai/team/)
4. [thehumanoid.ai/careers/](https://thehumanoid.ai/careers/)
5. [LinkedIn Company Page](https://www.linkedin.com/company/humanoidai)
6. [Jobs.ashbyhq.com/humanoid](https://jobs.ashbyhq.com/humanoid)

## Other roles at Humanoid

- [Senior Embedded Software Engineer - Build & Infrastructure](https://feeny.ai/job/senior-embedded-software-engineer-build-infrastructure-humanoid-us-boston-pmx6rnvttyyv) — US Boston, MA
- [Buyer (FTC)](https://feeny.ai/job/buyer-ftc-humanoid-london-380ty1qf7rkz) — London, United Kingdom
- [AI Data Collector](https://feeny.ai/job/ai-data-collector-humanoid-london-c2qc1rfsbf4x) — London, United Kingdom
- [Head of Investor Relations](https://feeny.ai/job/head-of-investor-relations-humanoid-london-d7qzv9arw0wk) — London, United Kingdom
- [Global Director of Treasury](https://feeny.ai/job/global-director-of-treasury-humanoid-london-5szpvajax22s) — London, United Kingdom
- [Finance Operations Analyst](https://feeny.ai/job/finance-operations-analyst-humanoid-london-2rmzqhhyrwxr) — London, United Kingdom
- [Senior Group Accountant](https://feeny.ai/job/senior-group-accountant-humanoid-london-cqasb4hw8wxf) — London, United Kingdom
- [VP of Sales](https://feeny.ai/job/vp-of-sales-humanoid-london-4avf2wah5q90) — London, United Kingdom
- [VP of Delivery](https://feeny.ai/job/vp-of-delivery-humanoid-london-jak0zreygwpy) — London, United Kingdom
- [Site Supply Chain Manager](https://feeny.ai/job/site-supply-chain-manager-humanoid-london-5hzphgkmrvx0) — London, United Kingdom