Senior ML Engineer Foundation Models & CUDA
TechTree's client is hiring a Senior ML Engineer to help build and scale a novel foundation model for automated software delivery in embedded systems.
This is a deeply technical role for an engineer who has built large-scale foundation models, developed custom CUDA kernels, and scaled distributed ML systems in production.
- Location: London - On-site
- Employment: Full-time
- Estimated compensation: £100,000-£150,000/year + equity
- Level: Mid-Senior / Senior
What you'll own
- Lead development and production deployment of a large-scale foundation model
- Design and implement custom CUDA kernels for performance-critical workloads
- Optimise training and inference across GPUs and distributed infrastructure
- Architect scalable ML systems with demanding reliability and performance requirements
- Profile and improve data pipelines, training loops, inference, and serving infrastructure
- Evaluate modern architectures including Mixture-of-Experts and state-space models
- Build internal tooling, benchmarks, evaluation systems, and observability infrastructure
- Work closely with founders on technical strategy and optimisation priorities
- Improve model quality, latency, scalability, and compute efficiency
What we're looking for
- Deep professional experience with CUDA C/C++
- Strong Python skills
- Experience building and shipping large-scale foundation models
- Proven custom CUDA kernel developmentExperience with distributed training and inference
- Strong knowledge of modern deep learning architectures
- Production expertise with PyTorch or another major ML framework
- Experience operating ML workloads on AWS, Azure, or GCP
- Strong understanding of ML observability, evaluation, and production SLOs
- Experience delivering complex systems in fast-moving environments
Strong advantages
- Built a foundation model from 01
- Early-stage AI startup or top-tier research lab experience
- Experience with MoE, state-space models, and advanced inference optimisation
- Built tooling that significantly improved engineering or research productivity
- Experience optimising large-scale GPU workloads for cost and performance
Why consider this opportunity?
- Work on a genuinely novel foundation-model problem
- Significant technical ownership from day one
- Direct influence over architecture and ML strategy
- Work closely with an experienced founding team
- Competitive compensation plus equity