Led · Series B · Oct 2021

H3 Dynamics is a global advanced air mobility technology company focused on decarbonizing air mobility through hydrogen-powered aircraft and drones. Its offerings include complete hydrogen drones, integrated aerial fuel-cell propulsion and refueling units, and a unified cloud SaaS that automates drone hardware and software to deliver autonomous outcomes as a service. The company sells inspection, performance monitoring, and live incident-response services across sectors and geographies (including the US, Canada, Latin America, France, Singapore, Japan, Australia, and Hong Kong). H3 Dynamics develops larger hydrogen systems and works on hydrogen aircraft integration and test flights in Toulouse, France, while producing and shipping complete hydrogen drones from Austin, Texas. The business intends to expand engineering and sales teams in Austin and Toulouse to scale production and integration work. Financially, the company closed a $26.0M Series B in October 2021 to accelerate its global mission.

Led · Equity · Nov 2018

Guardian Optical Technologies develops an in-car platform that pairs specially designed sensors capturing high-resolution video and three-dimensional depth maps with computer vision algorithms to monitor passengers and objects. The system detects when riders board, can identify the door used, estimate physical dimensions for tasks like child monitoring and adaptive airbag deployment, and can classify people and objects without direct line of sight. Engineering efforts prioritized cost reduction—Guardian says the processor is the most expensive component but that the solution is cost-effective versus comparable options and could save the auto industry $200 million a year by eliminating some special sensors. The company is actively prototyping the platform and intends to expand deployments into buses, trains, and other commercial and public transportation to provide real-time occupancy maps and flag behaviors such as vandalism and violence. Founded in 2015 and based in Tel Aviv, the startup will use recent funding to hire staff, better support ongoing customer projects, and increase its development pace. The article notes limited direct competition in the in-cabin computer vision space, citing products like Owl and Nauto that target different monitoring use cases. Guardian Optical builds a camera sensor that records high-resolution video and three-dimensional depth maps simultaneously and can detect motion as faint as a micrometer. Its machine-learning algorithms estimate passenger size to enable use cases such as adjusting airbag deployment force and monitoring young children in backseats. The sensor can detect when riders board, pinpoint the door they used, and classify people and objects inside the cabin without direct line of sight, which the company says is useful for passenger-aware and self-driving vehicles. Guardian Optical emphasizes cost-effectiveness, saying the processor is the most expensive component and projecting the solution could save the auto industry about $200 million a year by replacing other special sensors. The company plans to expand into buses, trains, and other commercial and public transportation for real-time occupancy maps and behavior flagging. Founded in 2015 and based in Tel Aviv, it previously raised $5.1 million in December 2017 and has announced a new $3.1 million funding round. Guardian Optical Technologies has developed a stand-alone automotive sensor that combines 2D video image recognition, 3D depth-mapping, and micro- to macro-motion detection to monitor an entire vehicle cabin with a single device. The system can detect heartbeat-level micro vibrations, locate and size occupants, distinguish people from objects, and identify objects even without direct line of sight. Built from low-cost, automotive-grade, off-the-shelf components, the company says the solution reduces the number of in-vehicle sensors and can lower costs by an estimated $20 per car (a potential $200 million annual saving for some manufacturers). Guardian's platform uses machine learning and real-time "big data" and image analysis and is designed to integrate with seatbelts, airbags, and other safety systems. The company positions the technology as forward-looking for autonomous vehicles and plans to use new funding to accelerate and expand technological development. Guardian is based in Tel Aviv, was founded in 2014, has patents pending and has won accelerator competitions sponsored by EcoMotion and Microsoft; it recently raised a $5.1M Series A.

Led · Series B · Oct 2018

AnyMind offers a suite of products for brands and online operators, including conversational commerce AnyChat, e-commerce management AnyX, manufacturing AnyFactory and logistics AnyLogi. The company integrates with marketplaces such as Shopee, Rakuten, Lazada, Amazon and Shopify and supports more than 1,000 brands, 1,110+ publishers and 1,400+ creators. AnyMind operates 17 offices across 13 markets with over 1,000 employees and reports $174 million in revenue. It grew through aggressive acquisitions (POKKT, Moindy, Acqua Media, LYFT and ENGAWA) and has retained many acquired founders in leadership roles. Management says it plans to use the latest funding to make additional acquisitions in Japan and globally. AnyMind began as an internet advertising business (formerly AdAsia) and has expanded into HR and broader marketing services, plus a new creators network. The company is moving into outdoor advertising in Thailand via a joint venture with VGI and has acquired Thai YouTube media company Moindy. It recently launched CastingAsia Creators Network to complement its marketing and media units. AnyMind says it has been profitable since early 2017; reported 2017 revenue was $26M (up from $12.9M the prior year) and the CEO expected revenue for the current year to be more than double 2017. The business operates across 11 Asian markets, with 12 offices (including a product development center in Vietnam), about 330 staff and more than 1,000 clients across three businesses. The company also maintains a sizeable office in Bangkok as it pursues regional expansion. AnyMind Group is a Singapore-based holding group (formed this year) that applies AI across online advertising, marketing, influencer and HR products, including AdAsia, TalentMind and CastingAsia. AdAsia offers ad solutions to publishers and advertisers, TalentMind focuses on HR, and CastingAsia handles influencer marketing. The company has been profitable since early 2017 and reported $26 million in revenue for 2017, up from $12.9 million in 2016, and expected 2018 revenue to be more than double 2017. AnyMind operates in 11 Asian markets with 12 offices (including a product development center in Vietnam), roughly 330 staff and serves over 1,000 clients; Southeast Asia supplies the majority of revenue while its one-year Japan operation accounts for about 30 percent. The company plans to expand market share across Asia, pursue strategic acquisitions, enter new markets such as India and the Middle East next year, and build additional software based on its AI technology. Leadership has indicated an intention to pursue an IPO around 2020, with potential destinations including Tokyo, New York or Hong Kong.

Participated · Equity · Sep 2018

Slamcore develops spatial intelligence software that uses stereo cameras and proprietary visual AI to continuously track the position and behavior of industrial vehicles without GPS, beacons, floor markers or other infrastructure. Its products, Slamcore Aware and Slamcore Alert, provide facility-wide vehicle visibility and monitor driver behavior and proximity to pedestrians and structures. The company says its solutions are deployed across more than 30 facilities in Europe and North America and have scaled to hundreds of units in under two years. Headquartered in London, Slamcore has raised a total of $40 million following the latest $14 million round. Company leadership and investors emphasize the operational datasets produced by deployments as a foundation for future physical AI applications.

Participated · Series C · Sep 2018

Blaize provides a purpose‑built, full‑stack programmable processor architecture and a low‑code/no‑code software platform (Blaize AI Studio and Blaize Picasso SDK) for edge and data‑center AI. Its solutions target automotive, mobility, retail, security, industrial automation, healthcare and other markets, emphasizing low power consumption, minimal latency and high efficiency. The company’s architecture and software aim to enable rapid development and deployment of applications with a highly optimized total cost of ownership. Blaize says current and next‑generation offerings will support computer vision, transformers and multimodal generative AI, and are designed to scale from the edge to the data center. The company has a comprehensive partner ecosystem and more than 200 employees across offices in El Dorado Hills (CA), San Jose (CA), Cary (NC), Hyderabad (India), Leeds and Kings Langley (UK), and Abu Dhabi (UAE). Blaize previously raised $224 million from strategic and financial investors and recently completed a $106 million financing to strengthen its balance sheet and accelerate its roadmap. In December 2023 Blaize announced its intention to go public via a business combination with Nasdaq‑listed BurTech Acquisition Corp. Blaize develops AI System-on-Chips (SoCs) and the Blaize AI Studio end-to-end application lifecycle software platform for edge and automotive computing. Its GSP architecture underpins first-generation edge hardware released in 2020 and a multi-year pipeline of products. The company targets automotive (in-cabin, out-of-vehicle, and autonomous functions), smart retail, smart city/metro and industrial markets. Blaize emphasizes power-efficient performance advantages for electric vehicles and centralized automotive compute, positioning its SoCs to improve efficiency and lower cost. The company will use the $71M Series D to accelerate its product roadmap and develop next-generation hardware and software to meet demand for higher performance, lower power and lower cost AI solutions. Prior to the Series D Blaize had raised US$87M from strategic and venture backers and employs 300+ people with headquarters in El Dorado Hills, California and teams/subsidiaries in Campbell and Cary (US), Hyderabad, Manila and the UK. ThinCI is an Artificial Intelligence computing architecture startup developing computing platforms for the expanding AI/machine learning market. Its core product is the Graph Streaming Processor (GSP) and a Software Development Kit (SDK), a general purpose fully programmable computing architecture. The GSP and SDK are designed to accelerate and optimize machine learning, deep learning, neural networks and computer vision workloads. ThinCI targets deployment across data center servers, edge infrastructure platforms and client platforms, from massive machine learning farms to sensor fusion and advanced neural networks for autonomous driving. The company said it will use proceeds from the Series C to continue to expand operations. ThinCI is led by CEO Dinakar Munagala and has teams in California, Hyderabad, India and the UK. Thinci develops machine-learning technology, including silicon and software, to enable embedded deep learning and vision processing in electronic systems. Led by CEO Dinakar Munagala, the company targets automotive applications such as advanced driver assistance systems and autonomous driving. Its technology can also be integrated into personal electronics and smart home automation systems. Thinci's product focus is on embedding vision processing into vehicle electronics to improve thermal system efficiency and optimize powertrain productivity. The article does not disclose revenue, user metrics, or the financial terms of the investment.