Piris Labs, a San Francisco-based startup developing photonic networking technology for artificial intelligence infrastructure, has raised $500,000 in seed funding to accelerate the commercialization of its optical interconnect platform. The financing will support product development, engineering expansion, and continued work on software and hardware designed to eliminate networking bottlenecks that limit the performance of modern AI data centers.

The seed round included backing from Untapped Ventures, Y Combinator, Batch Ventures, and Agile Angel. The investment follows the company’s acceptance into Y Combinator’s Winter 2026 accelerator program and reflects growing investor interest in next-generation infrastructure technologies designed to improve the efficiency of AI computing.

Founded in 2025 by Ali Khalatpour and Keyvan Moghadam, Piris Labs is focused on solving one of the most significant technical challenges facing artificial intelligence systems: the movement of data between processors. As AI models become larger and more computationally demanding, conventional copper-based networking has become a limiting factor, creating latency, increasing power consumption, and restricting scalability across data centers.

Piris Labs is developing a vertically integrated platform that combines proprietary photonic interconnect hardware with a purpose-built software stack. Instead of relying on traditional electrical connections, the company’s optical networking technology uses light to move information between computing resources. According to the company, this architecture is designed to deliver significantly lower latency, lower power consumption, and lower inference costs for AI workloads.

The company’s technology targets the growing demand for high-performance infrastructure supporting large language models, agentic AI systems, and other advanced inference applications. As hyperscale cloud providers continue deploying increasingly larger GPU clusters, efficient communication between processors has become just as important as improvements in chip performance itself.

Piris Labs says its optical interconnect platform is intended to transform the data center into what it describes as a “single node,” allowing distributed computing resources to communicate with substantially greater efficiency than conventional networking architectures. The company is also developing optical PCIe and memory disaggregation technologies that could allow AI systems to access shared memory resources across multiple computing nodes while reducing bottlenecks associated with existing infrastructure.

The newly raised capital will help expand the company’s engineering capabilities while accelerating product validation and customer engagement. Piris Labs is continuing to build both its hardware platform and software infrastructure simultaneously, enabling cloud providers and enterprise customers to integrate its technology into future AI deployments with minimal operational disruption.

The founding team combines expertise in photonics, artificial intelligence, and large-scale computing systems. Chief Executive Officer Ali Khalatpour previously conducted advanced photonics research at MIT, NASA, Stanford, and Harvard before founding the company. Co-founder and President Keyvan Moghadam previously held AI leadership roles at Meta and X, where he worked on building and scaling machine learning infrastructure. Their combined experience has shaped the company’s strategy of addressing both the physical networking layer and the software systems required for efficient AI execution.

Investor support from Untapped Ventures, Y Combinator, Batch Ventures, and Agile Angel highlights growing confidence in startups building foundational infrastructure for artificial intelligence rather than focusing solely on AI applications. As the demand for generative AI services continues to increase, infrastructure providers capable of improving computational efficiency are attracting increasing attention from venture investors.

Since its launch, Piris Labs has continued expanding its technology portfolio. The company has announced developments including a breakthrough optical conversion engine, the establishment of an advanced RF characterization laboratory, and its acceptance into the NVIDIA Inception program to support further development of its AI infrastructure platform. It has also outlined a long-term vision of enabling faster, lower-cost AI inference through optical networking technologies that overcome the limitations of traditional copper interconnects.

The broader AI infrastructure market continues to experience rapid growth as enterprises, cloud providers, and research organizations invest heavily in next-generation computing platforms. While advances in AI chips have received significant attention, networking and data movement have emerged as critical performance constraints for large-scale AI systems. Technologies that reduce latency and power consumption while increasing bandwidth are becoming increasingly important as organizations deploy larger and more sophisticated AI models.

With its $500,000 seed financing, Piris Labs plans to continue advancing its photonic networking platform and expand its presence within the rapidly growing AI infrastructure ecosystem. Backed by Untapped Ventures, Y Combinator, Batch Ventures, and Agile Angel, the company aims to build the networking layer that powers the next generation of scalable, energy-efficient artificial intelligence systems.

Related Articles