About Wayve and the team

Wayve is building embodied AI for the physical world, starting with autonomous driving. Instead of the hand-engineered, modular stacks that defined the first era of self-driving, we pioneered AV2.0: a single, end-to-end neural network that learns to drive from raw sensor data and generalizes to new cities, vehicles, and conditions. Our foundation models, the GAIA family of generative world models and the LINGO family of vision-language-action models, allow vehicles to perceive, reason, and act in the open world. We have driven zero-shot across hundreds of cities on three continents, and we are now scaling from proving the science to deploying it with leading automakers and mobility partners, including Nissan, Stellantis, and Uber.

This role sits in the AI Platform organization, on the data flywheel that powers every model we ship. Applied Scientists and ML Engineers on the team push the frontier on data curation, enrichment, foundation-model evaluation, and the models themselves. This role builds the platform underneath all of it: the pipelines, infrastructure, and systems that turn world-scale fleet data into high-signal training data, evaluate and train foundation models, and enable every team to run these workflows themselves. As deployment scales, the leverage is enormous: the better the platform, the faster the whole flywheel turns.

The role

We are hiring a senior Software Engineer to build the platform that powers Wayve’s data flywheel and foundation-model stack. This is the engineering counterpart to our Applied Scientist and ML Engineer roles: you build the systems they, and the wider company, depend on. It is high-leverage, high-visibility work with a clear path to deep system ownership.

• Build the systems that allow teams to turn world-scale driving data into high-signal training data, and evaluate and train foundation models on it.
• Replace ad-hoc scripts and manual handoffs with self-serve, observable products used across Science, Autonomy, and Evaluation.
• Every model Wayve ships runs on this platform: your work compounds across the entire fleet and roadmap.
• Work shoulder to shoulder with a world-class science and engineering team, with real deployment at global OEM scale (Nissan, Stellantis, Uber).
• TC3 / TC4 ownership of platform and infrastructure, with room to set technical direction as the platform matures.

What you will do

• Build and scale the data curation and enrichment pipelines that turn world-scale fleet data into high-signal training data: mining and active-learning loops, running model-based enrichments over billions of rows, and ensuring data quality at scale.
• Build the evaluation infrastructure behind foundation-model progress: harnesses for offline and closed-loop evaluation, metric and benchmark pipelines, and world-model-based evaluation.
• Build and optimize training and serving infrastructure for large pretrained models: distributed training, batched inference, and large-scale model backfills.
• Build the data-platform backbone: distributed data processing (Ray Data, Daft, Spark / Databricks), embedding and vector search (turbopuffer, Milvus), lakehouse formats (Lance, Iceberg), dataset versioning, and the enrichment and annotation catalog.
• Make it self-serve and reliable: turn one-off processes into products that other teams operate themselves, and own testing, observability, and on-call for what you ship.
• Partner closely with Applied Scientists and ML Engineers to take research from prototype to production at scale.

What we are looking for

• Strong production software engineering, especially production Python (services, APIs, large-scale data processing), and comfort owning and extending large codebases.
• Large-scale data and distributed-systems experience: batch and streaming pipelines, workflow orchestration (Flyte, Airflow, Dagster, or similar), and distributed processing (Spark / PySpark, Ray, Databricks, or equivalent).
• Systems design for scale: reliable, observable, high-throughput data or ML systems, with strong SQL and query and performance optimization.
• A track record of shipping and operating production systems that other teams depend on: testing, code review, observability, and on-call.
• Strong CS fundamentals and several years of production experience (roughly 6 or more for TC3, more for TC4), or equivalent; a degree in CS or comparable practical experience.
• Seniority to match the level: takes ambiguous, cross-team problems and drives them to completion, and at TC4 sets technical direction and multiplies the team.

Bonus points

• ML platform / MLOps: model registration, distributed training, and inference or serving optimization.
• Enough exposure to foundation models, world models, or ML evaluation to partner deeply with scientists.
• Embedding and vector search, annotation tooling, or feature and data catalogs.
• Kubernetes and modern data / lakehouse stacks (Databricks, Lance, Iceberg).
• Autonomous driving, robotics, or other large-scale sensor-data workflows.