21 hours ago
Oslo, NorwayEntry Level
Responsibilities
- Extend the LLM gateway with autorouting, caching, fallbacks, guardrails, telemetry, cost attribution, use-case onboarding, and access management.
- Deploy, operate, and benchmark self-hosted LLMs using vLLM or SGLang on Kubernetes in cloud and on-premise environments.
- Support approved on-device model serving with llama.cpp and evaluate models, engine settings, and GPU types for quality, throughput, latency, and cost.
- Improve observability, alerts, and per-team usage and cost reporting.
- Write and maintain evaluations for gateway and model changes and compare open-weight and proprietary models.
- Support engineers, data scientists, and journalists by answering questions and creating documentation, templates, and standards.
- Stay current with AI technologies and contribute to shared ML and AI knowledge across the company.
Requirements
- 0–2 years of experience as an ML engineer, AI engineer, software engineer, or similar, including relevant internships, thesis work, or serious open-source side projects.
- Master’s degree in Computer Science, Artificial Intelligence, or a related field.
- Strong Python skills and working knowledge of Docker, Kubernetes, and command-line tools.
- Experience running models locally and interest in LLM models, tools, recent releases, and GPU infrastructure.
- Comfort reading logs and metrics to diagnose failures.
- Clear communication skills for explaining technical solutions to non-technical colleagues.
- Curious, security-minded, collaborative, and adaptable approach to rapidly changing AI technologies.
- Hands-on LLM inference experience with vLLM, SGLang, llama.cpp, or similar is preferred.
- Knowledge of quantisation, distillation, caching strategies, and LLM or AI gateways is preferred.
- Experience with text, audio, and video data in a media context is preferred.
Benefits
- Inclusive and collaborative workplace with opportunities for continuous learning.
- Engineering culture based on trust, autonomy, and freedom to choose tools and influence ways of working.
- Support for specialization in MLOps pipelines, AI tools and frameworks, and LLM architectures.
- Work supporting public-interest journalism within Schibsted’s AI Foundations team in Sweden and Norway.
Tech Stack
Categories
About Schibsted
Schibsted is a Nordic media company that publishes news and runs digital products in Norway and Sweden, including VG, Aftonbladet, Aftenposten, and Svenska Dagbladet. It earns revenue from reader subscriptions, advertising, and branded content across web, apps, and print. Founded in 1839 and headquartered in Oslo, the group recently separated its marketplace operations into Vend.