Y Combinator-backed startup Photonium, which automates optical design systems with AI, was launched on Thursday, according to LinkedIn. The platform helps replace costly manual systems with end-to-end system optimization.
Founders Jennifer Song and Adam Mhatre came up with this startup after noting major issues with the design of optical systems. Optical systems are mission-critical across many high-growth sectors — including quantum technology, biomedical imaging, LiDAR for defense and autonomous vehicles, and metrology for semiconductor fabrication. However, their design involves companies pouring millions into teams of over 20 optical engineers with projects taking months or years to go from R&D to production. Tools are often fragmented and outdated, and optimization is manual and based on trial and error.
READ: AI startup Plexe debuts: Natural language meets autonomous agents for next-gen machine learning (May 28, 2025)
Engineers spend countless hours on repetitive workflows — constrained by tools that haven’t changed in decades. Photonium attempts to solve this by using AI to generate and optimize designs from product specs, predict tolerance impacts without slow simulations, recommend data-driven alignment and manufacturing tweaks, optimize for cost, size, and performance trade-offs, integrate with existing CAD workflows and optics simulation tools, and automate the full design-to-sourcing pipeline. Photonium works with complex optical setups across industries and handles the entire design process — from system generation and optimization to simulation, tolerancing, and sourcing.
The founders believe that optical systems are foundational to cutting-edge technologies from quantum computing to semiconductor fabrication; however, they are held back by outdated tools and fragmented workflows which slow innovation and inflate costs. Photonium aims to eliminate these bottlenecks and accelerate innovation across industries.
While Song has studied physics, math, and computer science at Harvard and built quantum optics experiments across Harvard, Stanford, and QuEra (a Google-backed quantum computing startup), Mhatre studied computational physics at Stanford, and worked on plasma simulation for fusion and large-scale astrophysical modeling of black holes. The duo claims to have firsthand experience dealing with inefficient engineering design, and in order to fix that, Photonium was founded.
“Very excited to share what Jennifer Song and I have been working on!” Mhatre wrote on LinkedIn.