Thesis: Cost-Aware Tiered Website Interaction with Calibrated Decision Models (AI)

Location: Västerås, Sweden; work is located at the Västerås office.

Company

hirify.global is a strategic partner in product and digital service development, combining cross-functional expertise to help clients develop solutions from idea to completion.

What you will do

• Design, implement, and evaluate a tiered website assistant for product searches, result filtering, order lookup, and review submission.
• Develop routing between lightweight decision models and progressively larger language models based on confidence.
• Compare the approach with LLM-only solutions using task success, token usage, operating cost, latency, and routing overhead.
• Conduct a literature review, prototype development, and experiments with representative website requests.
• Produce a working prototype, reproducible evaluation suite, and practical guidance for balancing cost, response time, and task success.

Requirements

• Master's student in Computer Science, Software Engineering, Artificial Intelligence, Machine Learning, or a related field; preferably applying as a pair.
• Good programming skills in Python or JavaScript/TypeScript and an interest in web applications and APIs.
• Foundational knowledge of machine learning and interest in LLMs, AI agents, and experimental evaluation.
• Ability to work independently, critically analyze results, and communicate findings clearly.
• Expected graduation in June 2027.
• Fluency in English, written and spoken.
• Fluency in Swedish, written and spoken, is normally required; candidates may have this assessed individually.

Nice to have

• Experience with Docker or model APIs.

Culture & Benefits

• Hands-on thesis work on a real project in a supportive and innovative environment.
• Regular feedback from byBrick Interface throughout the project.
• Start date is the beginning of January 2027.

Hiring process

• Apply through the careers page only; applications are accepted until 2026-11-30.
• Questions can be directed to Aldin Berisa and Edin Jelacic.