Hiring / looking for advice: precise visual SVG creation and editing with AI

I'm Katya, cofounder of BioRender, a "Canva for Science" platform that helps scientists create beautiful biological figures for research and communication. We're using AI to automate figure generation, but facing challenges with precision and scientific accuracy.

The issue we're running into is that LLMs cannot visually read and edit structured visuals (eg SVG or JSON files representing icons / shapes and coordinates).

Has anyone had to solve a similar problem and open to sharing notes? Or interested in joining our team to tackle this challenge together? Send me a note: katya@biorender.com

See example scientific figures here: https://app.biorender.com/biorender-templates