Master Thesis: Requirements to Verified RTL Releases: A Multi-Agent Traceability Framework for ASIC

Join our Team

About this opportunity

This thesis work proposes to develop and test a ‘Multi-agent ecosystem for feature-centric traceability and internal + external releases quality assurance and quality gating in for a cross-functional teams-oriented Organization supporting agile ways of development a complex ASIC IP.

What you will do

The proposed eco-system will ingest requirement specifications and connect them with Jira work items movement, RTL revisions for optimization or feature-changes, Documentation updates e.g. DS, VS, reference-model implementations, verification evidence, FPGA builds and release baselines.

Specialized agents are proposed to operate under clearly defined roles, while a coordinating agent will construct an auditable artifact graph, perform change-impact analysis and generate explainable quality-marker reports.

The research will combine LLM-based semantic reasoning with deterministic version, dependency and quality-gate rules. The prototype will be evaluated using traceability accuracy, warning precision, evidence completeness, unsupported-claim rate and practitioner assessment.

The expected contribution is a technically rigorous approach for improving XFT autonomy while preserving system-level consistency, release transparency and engineering accountability.

The skills you bring

• Curious mind
• Knowledge in Computer Science, Electrical Engineering, or a related field.
• Experience in at least one programming language (Python not mandatory).
• Eagerness to learn and explore new concepts.
• Bonus: experience in machine learning.