---
title: 'Research Scientist - Robot Learning (VLA / WAM) at SpAItial'
canonical: 'https://feeny.ai/job/research-scientist-robot-learning-vla-wam-spaitial-london-97ghdtvn8cmv'
type: 'job'
last_seen: '2026-09-13'
---

# Research Scientist - Robot Learning (VLA / WAM) at SpAItial

- **Company:** SpAItial
- **Location:** London, United Kingdom
- **Employment:** full-time
- **Work type:** onsite
- **Posted:** 2026-08-20
- **Last confirmed live:** 2026-09-13
- **Apply:** https://jobs.ashbyhq.com/spaitial/d20af35b-eb1c-49e5-a155-72d771fae4a7

## Job description

SpAItial is pioneering the next generation of World Models, pushing the boundaries of generative AI, computer vision, and the simulation of reality. We are moving beyond 2D pixels to build models that natively understand the physics and geometry of our world. Our mission is to redefine how industries, from robotics and AR/VR to gaming and cinema, generate and interact with physically-grounded 3D environments. We're seeking a Research Scientist to train the policies that turn a world model into a robot that acts. You will own vision-language-action (VLA) and world-action models (WAM) end to end, starting, including data, backbone, action representation, training runs, and the evaluation that tells us whether a policy is genuinely competent or merely lucky. A world model that understands geometry and physics still doesn't act on its own; the policy is what closes that gap. This is a senior, hands-on research role for someone who has already trained manipulation policies that worked, and who can say precisely why the ones that didn't failed.

## Responsibilities

- Own the training pipeline for vision-language-action (VLA) and world-action models (WAM) end to end, from data to a policy running on a robot.
- Contribute to setting the technical direction for embodied research at SpAItial.
- Close the sim-to-real gap through domain randomization, system identification, and calibration, and build evaluation that predicts real-world transfer.
- Adapt VLM backbones for control: encoder choice and adapter strategies, co-training.
- Curate and weight the training mix across heterogeneous robot datasets, spanning differing embodiments, action spaces, and sensor setups.
- Design action representation and decoding, including tokenization, chunking, diffusion, and flow-matching action experts.
- Build the world-model components that predict future observations conditioned on action.
- Run post-training: supervised fine-tuning onto target embodiments, and RL for robustness beyond demonstrations.

## Key Qualifications

- A PhD in robotics, machine learning, or computer vision with a robot learning focus, from the PhD alone or followed by industry experience.
- Publications at top venues such as (CoRL, RSS, ICRA, IROS or CVPR, ICCV, ECCV, NeurIPS), open-source work, and/or deployed systems.
- Deep experience with modern robot policy designs (VLA, WAM, diffusion), trained end to end rather than fine-tuned from a released checkpoint.
- Strong imitation learning fundamentals, and familiarity with RL fine-tuning of pretrained policies.
- Fluency with VLM backbones and how to adapt them for control.
- Expert Python and PyTorch, with multi-node distributed training experience (FSDP or equivalent).

At SpAItial, we are committed to creating a diverse and inclusive workplace. We welcome applications from people of all backgrounds, experiences, and perspectives. We are an equal opportunity employer and ensure all candidates are treated fairly throughout the recruitment process.

## About SpAItial

## Company Overview
- **One-liner**: SpAItial builds physically-grounded world models — AI that generates and reasons about the appearance and physics of real and imagined 3D environments.
- **Entity Type**: Private (Seed stage)
- **Headquarters**: London, United Kingdom (also has offices in Germany, United States, Luxembourg)
- **Founded**: 2024 (LinkedIn) / 2025-01-01 (company schema) — likely incorporated in late 2024, publicly launched May 2025
- **Founders**: Matthias Niessner (CEO), Luke Rogers (COO), Ricardo Martin‑Brualla, David Novotný

## Core Business
- **Primary industry/industries**: Spatial AI, 3D generative AI, foundation models, computer graphics
- **Target customers**: Developers, game studios, film/VFX, VR/AR platforms, robotics companies, industrial simulation (B2B); also provides a consumer app for world generation (B2C)
- **Mission or purpose**: “Bridging the virtual and physical – building physics‑consistent, spatio‑temporally grounded AI that understands the 3D world like humans do.”

## Products & Services
- **[Echo (Model Family)](https://spaitial.ai/)**: A series of Spatial Foundation Models (Echo‑1, Echo‑2, Echo HQ) that generate persistent 3D Gaussian Splat worlds from a single image, text prompt, or 360° panorama. Outputs are real‑time explorable and physically plausible.
- **[SpAItial API](https://spaitial.ai/)**: Programmatic access to Echo for agents, tools, simulations, and creative pipelines. Supports async generation, webhooks, and multiple output formats (SPZ, SOG, PLY, collision meshes).
- **[SpAItial App](https://spaitial.ai/)**: A web and desktop application for creating, editing, and sharing worlds generated by Echo. Includes image‑to‑world, text‑to‑world, sculpting, painting, and export tools.
- **Startup Program**: Credits, priority API access, early model access, and direct feedback loops for startups building with spatial AI.

## Market Standing
- **Valuation/Market Cap**: Not disclosed (private company)
- **Key Metric**: **Total Funding** — USD 13,000,000 (Seed round, closed June 27, 2025)
- **Notable Investors/Partners**:
  - Lead: Earlybird Venture Capital
  - Participants: Speedinvest
  - Notable angels: Robin Rombach (Black Forest Labs), Victor Riparbelli & Steffen Tjerrild (Synthesia), Christian Stiebner, Edward Grefenstette, Jon Barron, Rahul Garg, Sergej Epp (Sysdig/Palo Alto), Elias Schneider (Codesphere), Paul Whitehead (Zoopla), and others.
- **Growth Signals**:
  - Headcount: 20 employees (+2000% YoY from 1 employee the year prior)
  - Active job postings: 11 (monthly growth +22.2%, quarterly +175%)
  - Strong hiring across research, engineering, and ML infrastructure
  - Launched Echo‑2 in May 2026, Echo HQ in June 2026
  - Opened early access and startup program

## Competitive Advantages
- **Spatial Foundation Models (SFMs)**: Unlike LLMs, image, or video models, SFMs operate natively in 3D physical space, capturing geometry, materiality, and physics — enabling true spatial reasoning and physically consistent generation.
- **World‑class founding team**: Deep expertise from academia (TU Munich), big tech (Google, Meta), and scaled startups (Synthesia, Cazoo). Includes pioneers in generative 3D (GRAF, VoxGRAF, X‑Fields).
- **Real‑time, persistent 3D worlds**: Outputs are not just images or videos but fully explorable Gaussian Splat environments that can be edited, exported, and shared.
- **Strong investor backing**: $13M seed from top European VCs and influential angel investors validates the technology and vision.

## Strategic Focus
- **Model scaling**: Improving core model capabilities (Echo‑2, Echo HQ) for richer geometry, sharper detail, and stability.
- **Developer ecosystem**: Building out the API, documentation, and startup program to attract developers and partners.
- **Industry partnerships**: Piloting technology across gaming, film, CAD, VR/AR, and robotics — collaborating with leading organizations.
- **Talent acquisition**: Growing the team aggressively across research, engineering, and infrastructure.

## Why Work Here
- **Cutting‑edge research**: Work on frontier AI problems at the intersection of computer vision, graphics, and 3D generative models. Publishable research encouraged.
- **Strong team**: Colleagues from Meta, Google, TU Munich, Synthesia, and top labs. Low ego, high output.
- **Early stage impact**: Join at 20 people — your work directly shapes the product, culture, and direction of the company.
- **Global, hybrid‑friendly**: Headquarters in London with presence in Munich (Germany), New York (US), and Luxembourg. Mix of in‑person and remote collaboration.
- **Open roles** (as of mid‑2026): Research Scientist (3D Diffusion, World Models), Research Engineer (Graphics, World Models), ML Infrastructure Engineer — 11 active postings.
- **Perks**: Not publicly detailed, but seed‑stage startup typically offers equity, flexible time off, and direct access to founders.
- **Culture**: Described as “research meets real‑world application” — a blend of academic rigor and product‑focused execution.

## Sources
1. [spaitial.ai](https://spaitial.ai/) – Company homepage, product overview, API docs, blog
2. [spaitial.ai/about](https://spaitial.ai/about) – About page, mission, team, funding announcement
3. [spaitial.ai/careers](https://spaitial.ai/careers) – Careers page, open roles
4. [spaitial.ai/blog/announcing-spaitial](https://spaitial.ai/blog/announcing-spaitial) – Launch blog post with detailed team, funding, and vision
5. [LinkedIn – SpAItial AI](https://uk.linkedin.com/company/spaitial-ai) – Employee count, funding, headcount growth, open roles, talent sources

## Other roles at SpAItial

- [Technical Artist](https://feeny.ai/job/technical-artist-spaitial-london-kdgkjp9c2nx6) — London, United Kingdom
- [Business Operations Manager](https://feeny.ai/job/business-operations-manager-spaitial-london-fwsemcnhjvws) — London, United Kingdom
- [Research Scientist - 3D Reconstruction (SfM & SLAM)](https://feeny.ai/job/research-scientist-3d-reconstruction-sfm-slam-spaitial-london-dgx907pgrv08) — London, United Kingdom
- [Machine Learning Systems & Infrastructure Engineer](https://feeny.ai/job/machine-learning-systems-infrastructure-engineer-spaitial-london-yg2ny2hqnfkr) — London, United Kingdom
- [Machine Learning & Cloud Infra Engineer](https://feeny.ai/job/machine-learning-cloud-infra-engineer-spaitial-london-p5w3sxeysvq7) — London, United Kingdom
- [Research Scientist - 3D Diffusion](https://feeny.ai/job/research-scientist-3d-diffusion-spaitial-london-f21y5vjxxmzm) — London, United Kingdom
- [Research Engineer - 3D World Models](https://feeny.ai/job/research-engineer-3d-world-models-spaitial-london-68tajf2mpe9t) — London, United Kingdom
- [Research Engineer - Graphics](https://feeny.ai/job/research-engineer-graphics-spaitial-london-scfpn9d65rd2) — London, United Kingdom
- [Research Scientist - World Models](https://feeny.ai/job/research-scientist-world-models-spaitial-london-txvdt70j5zfj) — London, United Kingdom