Agentrys Raises $24.5M to Build AI Agents for Chip Design

Agentrys, a startup building AI agents for semiconductor engineering, raised $24.5 million in combined seed and pre-seed funding: a $19.1 million seed round led by Etna Labs, following an earlier $5.4 million pre-seed round led by chipmaker MediaTek, the company's first strategic backer.

What Agentrys Builds

Founded by Mark Ren, who brings nearly three decades of electronic design automation and AI research experience spanning NVIDIA Research and IBM Research, including leading ChipNeMo, one of the first industrial large language models built specifically for chip design, Agentrys is positioning itself in a category it calls Agentic Design Automation. The pitch: traditional electronic design automation tools handle individual tasks within chip design, while Agentrys aims to build intelligent systems that automate, learn from, and continuously improve entire engineering workflows, embedding an agentic workforce directly inside a chip design team's existing tools. The new capital will go toward recruiting, building out agent-native tooling, and expanding customer engagements across verification and physical design, two of the most labor-intensive and error-prone stages of the chip design process.

Why the Two-Stage Path Matters

MediaTek's early strategic involvement as first backer is notable: it signals that at least one major chipmaker sees enough value in agentic design automation to invest ahead of broader market validation, rather than simply becoming a customer later. The two-stage funding path, a strategic pre-seed followed by a larger institutional seed, is a pattern increasingly common among deep-tech startups selling into a small number of large, technically demanding customers, where an early strategic check functions as a credibility signal to the broader venture market. Ren's background leading ChipNeMo at NVIDIA gives Agentrys a technical pedigree that likely mattered as much to investors as any early revenue traction, particularly in a category where most generalist venture investors lack the domain expertise to independently evaluate the product's claims. See active deep-tech investors on AngelLinx for the broader landscape of funds backing companies in this category.

What This Means for Founders

Founders building in deep-tech, semiconductor tooling, or other technically specialized AI agent categories can use the investor match tool to identify investors with genuine domain expertise rather than generalist AI interest alone. Founders should be prepared to explain their burn rate and runway assumptions clearly when pitching technically complex products, since deep-tech investors typically scrutinize capital efficiency alongside technical differentiation.

See other deep-tech and AI tooling raises tracked in the AngelLinx newsroom. Founders ready to connect with active investors can explore AngelLinx @ angellinx.ai/register.

AngelLinx Intelligence | angellinx.ai