Master Thesis: Using Large Language Models and Ontologies to Improve Learning from Historical Safety Incidents (AI)

Location: Finspång, Östergötland County, Sweden

Company

hirify.global develops energy systems and technologies supporting reliable, sustainable energy generation and the energy transition.

What you will do

• Review safety representation approaches including Bow-Tie, FBS, STAMP/STPA, HFACS, energy-barrier models, and task-hazard-control models.
• Define an ontology or hybrid knowledge structure for millwright tasks, tools, components, hazards, exposures, injuries, and controls.
• Use a Large Language Model to extract and normalize safety knowledge from structured fields and free-text incident descriptions.
• Build a prototype ontology-based repository, knowledge graph, or discovery layer for retrieving relevant incidents, hazards, and controls.
• Compare ontology-based semantic retrieval with the current spreadsheet-based approach and evaluate extraction accuracy.
• Document methodology, limitations, and results, and communicate findings to technical and non-technical stakeholders.

Requirements

• Currently pursuing a master’s degree in Data Science, Machine Learning, Artificial Intelligence, Computer Science, Statistics, Knowledge Engineering, or a related field.
• Practical knowledge of machine learning, natural language processing, or Large Language Models.
• Interest in ontologies, knowledge graphs, semantic search, or structured information extraction.
• Experience with Python and common data science libraries.
• Analytical mindset and ability to work with incomplete, inconsistent, and ambiguous real-world text data.
• Ability to work autonomously while collaborating with safety specialists, domain experts, and data science stakeholders.
• Strong written and verbal communication skills, including the ability to document methods and present findings clearly.

Nice to have

• Familiarity with prompt design, structured outputs, or retrieval methods.

Culture & Benefits

• Full-time master’s thesis opportunity within the Advance Analytics team in hirify.global’s Digital Services organization.
• Collaboration with safety specialists, domain experts, and data science stakeholders.
• Opportunity to build a professional network for future cooperation within a global energy technology company.
• Exposure to central applications and platforms, solution architecture, service management, digital consulting, and project governance.
• Inclusive workplace with colleagues from more than 130 nationalities.

Hiring process

• Apply through the hirify.global careers portal using job ID 306309.
• Applications are accepted until 23 November 2026; selection is ongoing and the role may be filled earlier.