Master Thesis – Agentic AI for Stimulus Creation During Coverage Closure

Axis develops ARTPEC system-on-chip devices for its products, with ASIC verification playing an important role in hardware quality. While constrained-random testing covers much of a design, engineers often need to create targeted stimulus manually to reach the remaining coverage goals. This thesis explores whether an agentic AI pipeline can reduce that effort or turnaround time while generating valid, reproducible tests.

The assignment

• Review current approaches to agentic AI in ASIC verification.
• Benchmark the existing AI pipeline and reconstruct the historical manual effort for selected modules.
• Define the coverage scope following an initial analysis, likely prioritising functional coverage and including code coverage where relevant.
• Improve the pipeline's prompts, context selection, retrieval of code and examples, stimulus formats, and iterative feedback from tools.
• Evaluate the adapted pipeline on held-out coverage targets against both the initial pipeline and the original manual solution.
• Deliver an adapted pipeline, documented benchmark history, and an evidence-based assessment of its strengths and limitations. The focus is on the pipeline around the language model rather than training or fine-tuning model weights.

Who we are looking for

• Master's students in embedded systems, electrical engineering, computer science, or a relevant related field.
• A strong interest in embedded electronics and hardware development, particularly digital design, verification, and system-on-chip technology.
• Affiliation with a Swedish university or university college, either directly or through an exchange programme.
• Two students applying together as a pair.

What we offer

• A thesis project focused on agentic AI and ASIC verification in the context of Axis's chip development.
• An open, friendly, and collaborative company culture with an emphasis on developing skills and working methods.

Practical details

Location: Lund. Scope: two students, each completing a 30-credit master's thesis. Apply through the thesis advert by 27 November 2026; applications may be in Swedish or English. Submit separate applications, identify your thesis partner, and attach your CV and university transcript. Questions: Per Dagermo, Engineering Manager for the ASIC Verification team, +46 46 272 1800.