Participated · Series A · Apr 2026

Cognichip develops a domain-specific deep learning model intended to work alongside engineers to accelerate and reduce the cost of semiconductor chip design. The company says its approach—training on proprietary, licensed, and synthetic chip-design data—can cut development costs by more than 75% and shorten timelines by over half. Cognichip emerged from stealth last year after being founded in 2024 and has run demos, including a hackathon with San Jose State students designing RISC-V CPUs. It has been collaborating with undisclosed customers since September but has not yet produced a chip designed with its system. Financially, Cognichip has raised $93 million to date, including the recent $60 million round led by Seligman Ventures. The company is positioning itself against established EDA players like Synopsys and Cadence and newer startups such as ChipAgentsAI and Ricursive as it scales its model and commercial efforts.

Participated · Series B · Sep 2022

Movellus builds an Intelligent Clock Network™ platform and patented architectural innovations designed to maximize performance, power, and yield for advanced silicon architectures. Customers have integrated the platform into products spanning ultra-low-power edge AI devices to performance-centric cloud datacenter compute and AI offerings. The company says the new funding will be used to expand R&D and build out marketing and sales to support growth plans. Movellus closed a $23M Series B and reported a significant step up in valuation in the current round. SK hynix made a strategic investment and, along with other investors, endorsed Movellus’ technology and long-term strategy. The company is headquartered in San Jose with R&D centers in Michigan and Toronto. Movellus uses digital infrastructure to automatically produce design-optimized, process-portable analog blocks (PLLs, DLLs and LDOs) built from standard cells. Its initial products are optimized per design for markets ranging from edge AI to high-performance cloud computing. The company can create nano-watt PLLs for ultra-low-power, battery-less applications and serves customers across cloud computing, AI, edge AI and FPGA markets. Movellus aims to dramatically reduce SoC time-to-market while improving power, performance and area through automated analog design. The company has raised $10M to date, including a new $6M Series A. The new funding will be used to expand infrastructure, broaden the product portfolio, and increase market penetration. Movellus develops traditionally analog IPs using proprietary generators that leverage industry-standard digital implementation tools and standard cell libraries. Its generators automatically produce PLLs, DLLs and LDOs optimized per design and are process-portable. The generated IP delivers comparable performance with lower power consumption and a smaller footprint. IP from Movellus is customizable, silicon-verified, and can be delivered faster to meet aggressive schedules. Customers include semiconductor and systems companies in the artificial intelligence, networking, and FPGA segments. Movellus is located in San Jose, California. Movellus builds technology that enables digital design tools to automatically create and implement functions that were previously implemented using custom analog methods. Its flagship products are Movellus PLL, DLL and LDO Generators, which the company says deliver optimized, process‑portable, 100% digitally‑implemented IP in hours. Movellus claims the digital approach shortens implementation time, reduces power consumption, and yields a smaller footprint compared with traditional analog IP. The company targets semiconductor and systems customers in AI, networking, and FPGA segments. Movellus says the newly closed funding will be used to accelerate market deployment and adoption of its technology by semiconductor and system design companies. Movellus is located in San Jose, California.

Participated · Equity · Jun 2022

Medsien is a San Francisco, CA health technology company offering Remote Care Management solutions and services across the United States. The company’s core product set centers on remote care management and associated services. The article does not disclose founding year or operating metrics. Medsien announced a $4.3M funding round led by Merus Capital and Naples Technology Ventures, with participation from Candou Ventures. No additional financial instrument details or prior rounds are mentioned in the article.

Participated · Series A · Aug 2021

Bodo.ai builds a high-performance data processing engine that merges vectorized execution, high-performance computing technologies, and high-level compiler optimization to improve efficiency for Python and SQL workloads. The platform is positioned for demanding data engineering use cases and is now listed in the AWS Marketplace for easier procurement by AWS customers. Bodo.ai completed the AWS Foundational Technical Review, meeting security, reliability, and operational standards defined by the AWS Well-Architected Framework. The company has joined the Amazon Web Services Partner Network to expand its reach and collaboration with AWS and mutual customers. Bodo.ai plans to expand its technology into the government sector and support mission-critical applications following a new strategic investment. The company emphasizes reducing data infrastructure costs for customers and enabling scalable data processing for mission-critical workloads. Bodo.ai is developing a compiler and parallel compute platform to make Python portable and high‑performing across multiple hardware platforms for large-scale data and machine‑learning workloads. Its compiler automates parallelization so developers do not need new libraries, APIs, or to rewrite Python into other languages or GPU code. The technology is used for real-time data analytics across industries including financial services, telecommunications, retail and manufacturing. Bodo.ai was founded in 2019 by Behzad Nasre (CEO) and Ehsan Totoni (CTO) and is headquartered in San Francisco. With the new funding the company plans to triple its team and invest in R&D while building out marketing and sales. It is shifting from financing to a customer- and revenue-focused stage to drive broader adoption by the Python community.

Participated · Series A · Aug 2021

Bodo offers a parallel computing platform optimized for Python that aims to improve analytics and machine-learning workloads by compiling Python into fast binary executables via an “inferential compiler.” The software infers parallelism from application code and works with open-source libraries such as Pandas, NumPy and Numba, eliminating the need to rewrite Python into compiled languages. The company says its platform scales to more than 10,000 CPU cores, can process petabytes of data, improves analytics performance more than a hundredfold and boosts runtime efficiency tenfold. Bodo provides a free community edition for on-premises use and a cloud edition available on Amazon Web Services; pricing details for the cloud edition are on the company website. The firm is two years old and intends to build and sell its parallel computing platform to data-science and ML organizations. Financially, Bodo recently closed a new funding round to support those plans.