Junior ML Engineer (LLM Ops & Serving) in AI Foundations

Job description

Junior ML Engineer (LLM Ops & Serving) in AI Foundations About the team AI Foundations is Schibsted’s central AI platform team, responsible for core services and infrastructure that enable rapid and reliable ML and AI development. Located in Sweden and Norway, the team maintains platform components for ML orchestration; feature and vector stores; ML and LLM model serving; an LLM gateway; and supports the development of AI models and products across Schibsted. We collaborate with many other parts of the organisation to enable new AI initiatives, including experiment teams, product development, security & privacy, as well as journalists. We are now looking for a junior ML Engineer to strengthen the team. You’ll help improve our LLM gateway and model serving infrastructure, working with senior engineers on our team and with developers across the company. What will you do in this role? Together with senior engineers in the team, you'll help run and extend the platform that gives Schibsted's teams reliable, secure and cost-efficient access to LLMs. Your responsibilities include, but are not limited to: Extend our LLM gateway . Establish autorouting for cost-efficient model usage, add caching, fallbacks and guardrails, improve telemetry, and expand our cost attribution. Onboard new use cases, and manage access. Deploy and support self-hosted LLMs . Run vLLM- or SGLang-based models on Kubernetes, in the cloud and on-premise. Try out and benchmark new models, engine settings and GPU types for quality, throughput, latency and cost. Support approved golden paths for on-device model serving with e.g. llama.cpp. Make usage and cost visible . Extend observability solutions and alerts to support per-team cost reports so teams can see what they use, what it costs, and when something breaks. Guard quality with evals . Write and maintain evals that tell us whether a gateway or model change helped or hurt, and which open weight models perform as well as proprietary ones. Support our users . Be a first point of contact for engineers, data scientists and journalists building with AI, and turn recurring questions into docs, templates and standards. Explore! Contribute to a shared understanding of ML/AI knowledge across the company, challenging and inspiring the team by staying current with the bleeding edge of AI. What we’re looking for: You bring a strong engineering mindset, a structured approach to problem-solving, and the ability to collaborate effectively with cross-functional stakeholders. More specifically, you have: 0-2 years of experience as an ML engineer, AI engineer, software engineer or similar. We love to hear about internships, thesis work and serious side projects in the open source space too. Master's degree in Computer Science, Artificial Intelligence, or related field. You have an interest in LLM models, and all things that make a GPU go brrrr. You have experience running models locally, and have an opinion on which models are most relevant, tools, and recent releases across the LLM landscape. Strong Python skills, working knowledge of Docker, Kubernetes and the command line, and comfort reading logs and metrics to find out why something broke. A clear communication style: you can explain technical solutions to non-technical colleagues, including journalists. A curious and security-minded attitude to explore, learn, adapt, and share knowledge within the rapidly changing AI technology landscape. Nice to have Hands-on experience with LLM inference (vLLM, SGLang, llama.cpp, or similar). You can explain quantisation, distillation, and caching strategies. Familiarity with LLM/AI gateways. Experience working with text, audio, and video data in a media context. Our tech stack: Cloud infrastructure: Hybrid AWS/GCP, Kubernetes, Karpenter, KEDA CI/CD: Terraform, GitHub Actions Agentic solutions: AWS AgentCore, MCP, Pydantic AI, Promptfoo LLM solutions: LiteLLM, Bedrock, OpenWebUI Model serving: Triton Inference Server, vLLM Vector/Feature serving: Vespa, Feast Workload orchestration: Flyte, MLflow, Ray Monitoring & alerting: Grafana Coding: Python, Terraform, TypeScript Why we enjoy working here: At Schibsted, we value openness, curiosity, and collaboration. You’ll be part of a diverse, inclusive workplace that encourages bold thinking and continuous learning. Our work directly supports public interest journalism and we take that responsibility seriously. Our engineering culture is built on trust and autonomy. You’ll have significant freedom to choose your tools, influence how we work, and explore the technologies you find most compelling. Whether you're passionate about refining MLOps pipelines, diving into new AI tools and frameworks, or advancing LLM architectures, we support continuous learning and specialization. About the company Independent Journalism – That’s our business Schibsted Media Group includes some of the strongest media brands in the Nordics, including VG, Aftenposten, E24, Bergens Tidende, Stavanger Aftenblad, Aftonbladet, Svenska Dagbladet, Omni, and Podme. Every day, nearly seven million people turn to our editorial media to stay informed, engaged, and entertained through text, audio, images, and video. The trust of our users is crucial to us. To maintain this trust, we prioritise truth, verifiability, and transparency. Our 2,800 employees are based in Oslo, Bergen, Stavanger, Stockholm, Helsinki, Krakow, and Gdansk. We rely on all of them to succeed, through close collaboration across editorial teams, product and technology environments, and subscription and advertising units. What began as Christian Schibsted's small printing business in Christiania (now Oslo) in 1839 has grown into one of the leading media companies in the Nordics. For nearly two centuries, our journalism has empowered people, built communities, exposed abuses of power, and strengthened democracies. Our democracies depend on independent journalism. That’s our business.