Software Engineer - AI/LLM for Virtuoso

At Cadence, we hire and develop leaders and innovators who want to make an impact on the world of technology.

Cadence Design Systems is the leading provider of design automation tools for electronic and intelligent systems design. Virtuoso Development Engineer is responsible for designing, implementing, and maintaining software used in virtuoso platform in the epoch of emerging Agentic AI frameworks. Engineer will be part of a team focused on modernizing integrated circuit designing, simulation, and optimization tools to integrate into emerging Agentic AI frameworks in the broader electronic design automation context.

Responsibilities

• Design and implement new software features and interfaces in the Cadence Virtuoso platform.
• Contribute to continuous integration, unit testing, and code quality processes using modern C++ paradigms and AI-assisted development workflows.
• Collaborate with cross-functional engineering teams across time zones (US, China, EU, UK, and India) to deliver production-quality simulation technologies.

Required Qualifications

• MS in Computer Science/Computer Engineering.
• Proficient in Python, C++.
• Proficient in English (written and verbal) for cross-site collaboration.
• Candidate need to have the necessary computer science background to be able to design and deploy production software systems that make use of ML/LLM technology. You will not be developing new ML models themselves, but innovating to use AI technology in an EDA software framework and pushing the boundaries of Agentic AI within those frameworks.
• Desire to develop and deploy next-generation production EDA tools that have real-world impact on productivity of integrated circuit designers.

Skills of Interest
• Software quality practices: Experience with unit testing, code review processes.
• AI-assisted development workflows: Ability to integrate AI tools into engineering workflows to enhance productivity and code quality.
• Machine Learning Engineering: Basic background in machine learning, data pipelines, predictive modeling, and deployment. Practical knowledge of frontier LLMs (e.g., DeepSeek, Qwen, GPT/Claude families) and how choices affect latency, cost, and reliability.
• Agent architecture: Familiarity with concepts such as ReAct (reason-act loops), planning/evaluation/self-correction, and persistent memory design; frameworks (e.g., LangChain, Dify, Coze); MCP (Model Context Protocol), function/tool calling, and structured outputs.
• Data and retrieval: Understanding of RAG, retrieval pipelines, embeddings, chunking, and grounding.