Participated · Series A · Apr 2023

Mojo Vision pioneers a wafers-in, wafers-out micro-LED platform intended to enable massively parallel optical I/O for AI infrastructure. Its integrated approach combines advanced 300mm CMOS circuitry, GaN-on-silicon micro-LED emitters, silicon photodetectors, multicore fiber bundles, custom micro-lens arrays, and proprietary software and tools. The company says this architecture enables thousands of optical channels in a compact footprint, delivering high bandwidth density, low energy per bit, and built-in redundancy. Mojo Vision has framed its technology as relevant to terrestrial data centers as well as distributed and orbital computing architectures where power, weight, and scalability matter. The company is based in Cupertino, Calif., and recently completed a $75 million financing the prior August before securing the $17.5 million strategic investment from Future Ventures. Mojo Vision is collaborating with Marvell to develop high-density micro-LED connectivity solutions as it advances commercialization.

Participated · Series B · Jul 2022

DELFI Diagnostics is a Palo Alto–based developer of blood-based cancer detection tests that use fragmentomics and advanced machine-learning methods applied to whole-genome sequencing data. The platform analyzes patterns of cell-free DNA fragments to detect chaotic signatures left by cancer cells. DELFI has commercialized FirstLook Lung, a blood test that evaluates DNA fragment patterns to indicate the presence of lung cancer. In an independent validation, FirstLook Lung demonstrated 80% sensitivity in a screening population, including early-stage detection, and a 99.8% negative predictive value. The company received an equity investment from the Merck Global Health Innovation Fund and intends to use the funds to accelerate development and commercialization of its cancer detection solutions. Led by CEO Susan Tousi, DELFI positions its fragmentomics and AI technologies as a potential high-precision, population-scale approach to early cancer detection at a disruptive price. Delfi Diagnostics has built a platform that analyzes cell-free DNA fragments across the entire genome using a proprietary, scalable whole-genome fragmentation approach. The platform is designed to enable a single, low-cost, and straightforward lab process for multiple liquid-biopsy tests and leverages machine learning to deliver high-performing results. Led by CEO and founder Victor Velculescu, the company is validating its technology for early detection of lung and other cancers in a 15,000-person prospective trial called CASCADE-LUNG. Delfi is working with multiple research institutions to develop additional applications including screening for other cancer types, multi-cancer early detection, and treatment monitoring. The company intends to use the funds raised to continue development and commercialization of single-cancer early detection, multi-cancer early detection, and treatment-monitoring blood tests. Headquarters are listed in Palo Alto, CA and Baltimore, MD. Delfi Diagnostics is a Baltimore-based company developing a new class of high-performance, affordable liquid biopsy tests for early cancer detection. Its platform leverages machine learning and genome-wide fragmentation profiles of cell-free DNA to detect evidence of tumors from a blood sample and to pinpoint tumor location. The company is applying this platform to a range of products to detect specific cancers, such as lung cancer in at-risk individuals, and to identify multiple cancers in any individual. Delfi plans to use the funding to expand its team of cancer researchers and machine learning experts and to validate its next-generation liquid biopsy technology through multiple prospective clinical studies. Financially, Delfi raised $100M in a Series A financing led by OrbiMed. The company is led by founder and CEO Victor Velculescu.

Participated · Seed · Feb 2022

Neo Cybernetica is operating in stealth mode and developing scalable technologies in the field of cybernetics. Specific details about its product are not currently available in public reports. The company says it is leveraging breakthroughs in neuroscience, cognitive psychology, artificial intelligence, robotics, simulation, machine learning, and reinforcement learning. Neo Cybernetica plans to use new funding to scale R&D and accelerate technology development. The company is led by CEO Jeremy Achin and CTO Dmytro Zahanych. DataRobot board director Tony Florence is joining Neo Cybernetica’s board of directors.

Led · Series A · Apr 2020

HeartVista develops AI-assisted MRI software, centered on its One-Click Cardiac Package that enables real-time MRI and one-click cardiac localization. The One-Click software prescribes standard cardiac views in as few as 10 seconds while the patient breathes freely and uses a neural network to detect artifacts and warn clinicians of low-quality images. The company plans to build AI-guided musculoskeletal and neural packages, expand sales, marketing and regulatory operations worldwide, and deepen strategic partnerships with leading cardiology centers. HeartVista recently received its fifth FDA 510(k) clearance for its One-Click Cardiac Package, appointed Itamar Kandel as CEO, formed a Medical and Scientific Advisory Board, and expanded its Board of Directors. The company is offering free use of its software for the duration of the COVID-19 pandemic and will provide free remote cardiac scanning services to providers that lack staff. HeartVista’s technology originated at the Stanford Magnetic Resonance Systems Research Laboratory and the company is funded by Khosla Ventures and grants from the National Institutes of Health, including SBIR funding, with total funding now at $16.35 million.

Participated · Series C · Jul 2017

Cerebras Systems designs and sells what it describes as the fastest AI infrastructure in the world, centered on its Wafer-Scale Engine 3 (WSE-3) processor. The company states the WSE-3 is the largest commercialized AI processor, 58 times larger than the largest GPU, delivering inference and training up to 15 times faster while using a fraction of the power per unit compute. Cerebras offers solutions both on premises and in the cloud and reports deployments with leading corporations, research institutes, and governments across four continents. Financially, the company completed a $1 billion Series G in September 2025 and a $1 billion Series H in January 2026, and has now secured an $850 million revolving credit facility, bringing total capital raised to $2.85 billion in the last eight months. The company intends to use the new credit facility to expand data center capacity and support its growth plans.