Securitas Group

Securitas is a world-leading safety and security solutions partner that helps make your world a safer place. By leveraging technology in partnership with our clients, we offer a broad portfolio of value-enhancing services and solutions integrated across the security value chain – from on-site services to advanced monitoring, comprehensive risk prediction and advisory services.

With around 322 000 employees in 44 markets, our innovative, holistic approach with local and global expertise makes us a trusted business partner to many of the world’s best-known companies. Benefitting from almost nine decades of deep experience and guided by our values of integrity, vigilance, and helpfulness, we create sustainable value by helping our clients optimize their operations and protect what matters most - their people and assets.

AI Team at Securitas

At Securitas, our colleagues show up every day to help keep communities and organizations safe. Our job in the AI team is to make sure they're equipped with the best tools and intelligence possible.

We are Securitas' Specialized AI Team - the internal center of excellence for advanced, custom AI. We don't work on generic tools or off-the-shelf solutions. We build the AI capabilities that require deep technical and domain expertise, and that directly move the needle on how Securitas operates at scale.

About the role

You will be a key technical voice in a small, focused team - someone who shapes how we approach problems, not just solves them. You'll lead the design, development, and production deployment of ML and GenAI solutions that turn raw data into actionable intelligence across a range of real-world problems.

The stack we work with

Python · PyTorch · LLMs (OpenAI, Gemini, Claude, open-source) · RAG pipelines, Hugging Face · Python analytical tools (DuckDB, polars, Pandas, and more) · Streamlit & Dash · Claude Code · GitHub Copilot · React · Databricks · Docker · SQL/NoSQL · Azure/GCP

Responsibilities

• Owning the architecture of LLM-powered pipelines that extract structure and insight from large volumes of unstructured text, such as incident reports, operational logs, client data.

• Designing and stress-testing end-to-end GenAI architectures (e.g. RAG), prompt strategies, and evaluation frameworks - relevance, faithfulness, hallucination rates, latency tradeoffs - and setting the bar for what "good" looks like on the team.

• Building and productionizing workforce management models - demand forecasting, shift scheduling optimization, and attrition modeling - that help deploy officers more effectively.

• Developing client churn models that give the business early, actionable retention signals.

• Driving the end-to-end ML lifecycle: from problem framing and data strategy through to monitored, production-grade systems.

• Translating ambiguous business problems into concrete technical roadmaps - and pushing back when the framing is wrong.

• Mentoring junior data scientists and setting technical standards across the team.

• Presenting findings, model behavior, and tradeoffs to senior stakeholders clearly and credibly.

What you'll bring

Must-haves

• Around 5+ years of professional data science experience, with a clear track record of successes.

• Deep Python skills and strong software engineering habits - your code is readable, tested, and maintainable.

• Advanced NLP experience and hands-on work with LLMs at a level beyond prompt experimentation - fine-tuning, evaluation, deployment.

• A rigorous approach to GenAI evaluation: you've built frameworks to measure output quality, catch failure modes, and make principled tradeoffs with full lifecycle thinking.

• Experience with MLOps fundamentals: deployment, serving, and monitoring of models, CI/CD, Docker, application and service logging, and reproducible pipelines.

• Using modern AI coding tools to work as a highly productive data scientist - rapidly exploring ideas, writing and refactoring code, and debugging faster while keeping a critical eye on outputs.

• Strong analytical instincts - you can tell when a result is too good to be true and you know how to find out why.

• Communication skills sharp enough to run a stakeholder presentation and a code review on the same day.

• Experience with Databricks for large-scale data processing and collaborative workflows.

• Familiarity with LLM evaluation frameworks (Ragas, LangSmith, or similar).

• Hands-on experience with forecasting and optimization problems - scheduling, demand planning, or similar.

• Experience building interactive data tools for end users, including front-end design, authentication layers, and logging.

• Background in a domain where decisions have real operational consequences - logistics, healthcare, security, or similar.

Working conditions

The role is open for candidates based in Malmö or Stockholm (with preference for applicants in Malmö). It's a hybrid working model.

What we offer

At Securitas we believe in doing the right thing and doing it well. For our customers and our employees. Our employees come from all walks of life and bring with them many talents and perspectives. We aim for diverse representation throughout the company, and we are committed to equal pay, safe working conditions, gender balance and an inclusive work environment with a wide range of skills and development opportunities.

If this sounds like the right next step in your professional career, don't hesitate and apply!