Participated · Series A · Dec 2019

Teraki builds AI and edge data-processing software that makes applications and algorithms run up to 10X more efficiently on constrained automotive and IoT hardware while preserving data quality for machine learning. Its Intelligent Signal Processing product extracts more information from telematics, video and 3D point-cloud signals with an ultra-low footprint suitable for safety-critical and certifiable embedded systems. The technology is used for predictive maintenance, crash detection, object detection, sensor fusion, AD and ADAS applications, and the company cites engagements with automotive OEMs, Tier 1 suppliers and semiconductor customers. Teraki highlights partnerships with Microsoft and Infineon and says its software enables devices to communicate and accelerate AI up to 10X more efficiently. The company employs about 40 people and will use new funding to grow the team, accelerate the product roadmap and expand into the U.S. and Asia (it recently opened offices in Seoul and Tokyo). Financially, Teraki closed an $11M Series A, bringing total capital invested to $16.3M, and plans to target the $395 billion automotive electronics market while expanding into drones, robotics and other IoT sectors. Teraki develops embedded, pre-processing software that reduces and structures the vast amounts of sensor and control-unit data generated in vehicles so lower-power automotive hardware can process it. When embedded in ECUs, MCUs and TCUs, the software aims to improve vehicle safety and autonomy while lowering operational costs. The company has completed several pre-production validations with premium automotive manufacturers and has performed integrations on a variety of microcontrollers. Teraki plans to use new funding to continue to expand its operations. The company is led by CEO and co-founder Daniel Richart and co-founder/CTO Markus Kopf. As of this round the company has raised a total of $5.3M. Teraki provides AI-based Intelligent Signal Processing software that enables edge data processing and delivers more than a 10X increase in automotive chip, communications and learning performance. Its technology targets resource- and cost-constrained automotive ECUs and networks to make highly accurate AI applications possible at scale in embedded environments. Teraki has completed several pre-production validations with premium automotive manufacturers and chip suppliers and has ongoing proofs of concept with additional OEMs. The company plans to use recent financing to support growth of automotive customer implementations and to accelerate development of its product roadmap. Teraki is privately held, headquartered in Berlin, and was founded by co-founders Daniel Richart and Markus Kopf, who lead a research team of more than 10 researchers. The team emphasizes adapting high-end data-analytics techniques to the constrained automotive infrastructure and other data-intensive IoT markets over time.

Participated · Equity · May 2019

Zipline builds and operates autonomous fixed-wing drone systems to deliver packages, with a track record of medical supply deliveries in Africa and expanding consumer deliveries in the U.S. Its commercial roster includes partners such as Walmart and more than a dozen restaurant brands including Panera, Chipotle, and Wendy’s. The company operates across four continents and has completed about two million deliveries to date. Financially, Zipline closed an extended $800 million Series H earlier this year and was most recently valued at $7.6 billion. Under its new strategic partnership with Uber, Zipline will integrate with Uber Eats and begin deliveries in existing U.S. markets later this year, aiming to scale rapidly toward a target of one million daily deliveries by 2029. The deal is intended to accelerate mainstream adoption by pairing Zipline’s infrastructure and regulatory approvals in select markets with Uber’s on-demand marketplace.

Participated · Series A · Jan 2018

Knightscope is an advanced security technology company based in Silicon Valley that builds autonomous security robots called Autonomous Data Machines (ADMs). Its ADMs are designed to deter, detect and report threats autonomously and are deployed by clients across multiple states. The company says it presently holds contracts in 15 states operating across four time zones on a 24/7 basis. Knightscope entered a loan and security agreement with Silicon Valley Bank to support production and scaling of its robots as it expands nationwide. The company will use the funds drawn and available under the SVB facility to finance production of additional ADMs to meet client order demands. The press release includes forward-looking statements and notes risks that could affect projected business performance, including the possibility of operating losses. Knightscope develops advanced physical security technologies centered on autonomous security robots that deter, detect and report, helping clients reduce vehicle break‑ins, vandalism and trespassing. Led by chairman and CEO William Santana Li, the company focuses on enhancing U.S. security operations. The firm will use the new funding to continue to expand operations and broaden its business reach. Knightscope raised over $25M in the cited financing, adding to prior capital and strengthening its balance sheet. Since inception in 2013 the company has raised more than $39M total. The company is headquartered in Mountain View, California. Knightscope, founded in 2013 by William Santana Li, develops advanced physical security technologies from its Mountain View, Calif. base. The company combines software, hardware and human components to deliver solutions focused on improving security operations and enhanced anomaly detection. It is raising capital through a Reg A+ mini-IPO offering of Series m Preferred Stock. Konica Minolta has invested $1m immediately and committed an additional $2m upon achievement of certain milestones, for up to $3m in funding. The capital will be invested in the Series m Preferred Stock offering and interested investors may purchase preferred stock online at SeedInvest. Knightscope builds Autonomous Data Machines (ADMs), fully autonomous robots with onboard sensors designed to monitor surroundings and provide a physical security presence. A single ADM can generate over 90 terabytes of data per year and communicates via wireless networks with the Knightscope Security Operations Center (KSOC), a browser-based interface that issues real-time alerts. The company’s first machine, the K5, is being offered for preorder on a Machine‑as‑a‑Service basis to select Silicon Valley customers at $6.25 per hour for 24/7 deployments on one- to three-year terms. Knightscope’s platform combines robotics, predictive analytics, sensors and collaborative social engagement to predict and prevent crime and to free humans for higher‑value security tasks. The business plans to accelerate deployments in Silicon Valley and continue technology development using proceeds from its recent financing. Knightscope is based in Mountain View, California. Knightscope is a robotics startup based in Silicon Valley and Tokyo that builds autonomous robots designed to predict and prevent crime. The company frames its technology as a blend of security automation — the article described it as "Robocop meets the Roomba." Knightscope recently raised $1 million in a seed round from NTT DOCOMO Ventures and unnamed angel investors. With the new funding, the company plans to begin beta testing its products. The article did not disclose revenue, user metrics, prior financings, or a founding year. No additional financial terms or investor specifics beyond NTT DOCOMO Ventures and unnamed angels were provided.

Participated · Series A · May 2015

Wonder Workshop makes small, programmable robots that children can control via mobile devices to learn computer science and collaborate. Its robots are used in about 12,000 elementary schools and the company runs a robotics competition to teach STEM concepts. The startup recently launched Cue, a $200 coding robot aimed at ages 11 and up with multiple sensors and customizable avatars. Leadership says the company is expanding product focus beyond core instruction for 8–10-year-olds to offerings that better engage pre-teens and teens. Wonder Workshop announced plans to use new funding to expand its global footprint and invest in development of future devices. The company is Bay Area–based and led by CEO Vikas Gupta. Wonder Workshop makes the Dash and Dot robots and companion apps to teach coding and computer science to kids as young as six. The company’s products include advanced sensors, lights and sounds that turn CS principles into hands-on learning and play. Wonder Workshop reports use in more than 7,000 elementary schools (4,000 in the U.S. and 3,000 internationally) and distribution in six major U.S. and Canadian retailers, with children in over 37 countries using its products. The company announced a $20 million Series B financing to fund international expansion, with a particular focus on China and Asia, plus product investment and management growth. Recent hires include CFO Monica Hushen as the company scales operations. Wonder Workshop was founded in 2012 and is based in San Mateo, California. Wonder Workshop builds Dash & Dot smart robots and companion apps that let children learn coding concepts while exploring creative play. The company offers four free apps tailored to different developmental stages and emphasizes problem solving and curiosity. Founded in 2012 by Vikas Gupta, Saurabh Gupta and Mikal Greaves and based in San Mateo, CA, Wonder Workshop launched in December 2014. Since launch the company reports over $3.5M robots have been delivered to 37 countries, including the US, Canada, Australia, New Zealand, parts of Europe and several Asian markets. Financially, the company has raised $15.9M to date and recently secured an additional $6.9M. The new capital is earmarked to expand into new markets and to develop manufacturing and strategic business channels. Play-i builds Bo and Yana, educational interactive robots designed to teach children programming fundamentals via companion tablet and mobile apps. Founded in 2012 by Vikas Gupta, Saurabh Gupta and Mikal Graves, the company presents visual sequences of commands that introduce concepts like subroutines, functions and loops. Commands can be recorded in languages including Java and Python so users can increase complexity as they progress. Play-i has more than 11,000 pre-orders and has signed over 200 schools for pilot programs, and it is working with 100+ developers for early API access. The company is focused on improving its software and app interfaces for multiple age groups and plans a broad API release when it begins shipping Bo and Yana later this year. Play-i intends to pursue manufacturing and retail/eCommerce distribution partnerships and plans to double its 15-person team over the next year, prioritizing engineering hires. Play-i builds interactive, toy-like robots—Bo and Yana—controlled via a companion iPad app that presents visual sequences and simple commands to teach programming concepts through play, music and storytelling. The app records commands and can display them in programming languages such as Java and Python so learning scales with age. Founders include Vikas Gupta (former head of consumer payments at Google), Saurabh Gupta (led iPod software at Apple) and Mikal Greaves (led product design and manufacturing at Frog Design). The company has over 10,000 pre-orders and plans to ship the robots next summer; it will spend the next six months finalizing manufacturing and distribution partnerships. Play-i intends to sell through its website and via online and brick-and-mortar retailers, though retail deals are still in the works. The product enters an emerging education market alongside offerings like Cargo-Bot, Move the Turtle and Bee-Bot as STEM education demand grows.