6 months ago
Stockholm, SwedenMid Level
H1B Sponsor
Responsibilities
- Architect and optimize production AI workloads on Modal for customers including Suno, Lovable, Cognition, and Meta.
- Work on workloads including LLM serving, model training, audio pipelines, and scientific computing.
- Conduct technical demos, experiments, and proofs of concept that demonstrate Modal’s performance advantages.
- Collaborate with Modal’s product and sales teams as an engineer and product stakeholder.
- Build trusted relationships with CTOs, VPs of Engineering, and ML leads at frontier AI companies.
- Contribute to open-source projects such as SGLang and publish technical content about Modal’s capabilities across the AI stack.
Requirements
- At least 2 years of professional ML engineering experience, ideally involving inference optimization, model training, GPU programming, or ML infrastructure.
- Familiarity with ML serving or training toolchains such as vLLM, SGLang, slime, verl, and TRL, with the ability to go deep on at least one.
- Strong communication skills for discussing technical architecture and tradeoffs with engineering teams and technical leadership.
- Genuine interest in working directly with customers to understand and solve their problems.
- Side projects, open-source contributions, or published ML or systems-performance work are a bonus.
- Willingness to work in person in Stockholm.
Benefits
- In-person work in Stockholm
- Opportunities to contribute to open-source projects and publish technical content
- Collaboration with leading AI companies and foundation model labs
Tech Stack
Seaborn
Categories
Forward DeployedML Engineering
About Modal
Customers rely on Modal for instant GPU access, sub-second container starts, and native storage, so it's simple to serve low-latency inference, fine-tune models, and access production-ready sandboxes at scale. Every era of computing came with new workloads that previous infrastructure couldn't serve: mainframes, databases, the cloud. Each time, the company that rebuilt the layer underneath defined the decade. AI is no different, except it touches everything instead of one slice. The window to build is open right now.