The Venture Codex

Participated · Seed · Aug 2025

Kotoba Technologies, founded in 2023 by Noriyuki Kojima (CEO) and Jungo Kasai (CTO), builds the Koto family of real-time speech models — including speech-to-speech (S2S), ultra-low-latency speech-to-text (ASR), and text-to-speech (TTS) — optimized for Japanese, Korean, and Chinese. Koto models are deployable both in datacenters and on-device, including smartphones and wearables, and Kotoba reports outperforming leading models from American and European research labs on recognition and synthesis. The company is already in production with Fortune Global 500 organizations and AI-native startups, powering AI voice agents, contact-center voice interfaces, wearables, and simultaneous translation. Kotoba has released an alpha API and a Python SDK to broaden developer access and enable testing of its S2S and on-device models. Its consumer-facing Kotoba app delivers simultaneous translation, note-taking, and AI summaries across 21 languages and has surpassed 180,000 users. The company is headquartered in San Francisco with a Japan office in Tokyo and has raised $23 million in total funding to date.

Participated · Series B · Feb 2025

EnCharge is a semiconductor startup spun out from Princeton that builds analog memory chips and full‑stack AI accelerators aimed at running existing AI models at the edge (laptops, handsets, wearables). The company says its accelerators use about 20x less energy to run workloads compared with other chips on the market and has developed noise‑resilient analog designs and accompanying software. EnCharge expects to have its first chips on the market later this year and is working closely with TSMC to manufacture those devices. The startup focuses on inference rather than training and continues R&D with algorithms that could expand use cases. EnCharge has received non-dilutive support via grants from DARPA and the Department of Defense and counts strategic backers such as In-Q-Tel among its investors. The company is Santa Clara‑based and emerged from stealth in 2022 after nearly a decade of research at Princeton. EnCharge AI commercializes next-generation in-memory computing chips and full-stack AI accelerators that use switched-capacitor analog techniques to deliver large improvements in compute efficiency and density. The company says its technology targets decentralizing AI inference from data centers to phones, laptops, vehicles, and factories. EnCharge was launched in 2022 and is led by co-founder and CEO Dr. Naveen Verma, with Dr. Kailash Gopalakrishnan as CTO. Headquartered in Santa Clara, Calif., the company is building on years of Princeton lab work and prior DoD/DARPA-funded research. Financially, EnCharge has raised $45 million to date and recently announced a $22.6 million funding round to develop next-generation full-stack AI compute solutions. It is also participating with Princeton University in a DARPA OPTIMA-supported multi-year project funded by an $18.6 million grant. EnCharge AI builds custom plug-in hardware on a standard PCIe form factor that uses in-memory computing to accelerate AI workloads for servers and network edge machines. The company’s chips are designed around capacitors rather than transistors to improve robustness against voltage and temperature fluctuations. EnCharge is developing accompanying software to enable customers to adapt different neural networks to the custom hardware and to keep the solution scalable. The firm emerged from research and DARPA- and federally funded R&D that Verma led beginning in 2017 and commercializes those test chips and architectures. EnCharge has roughly 25 employees, is based in Santa Clara, and is pre-revenue. Proceeds from the recent raise will be used for hardware and software development and to support new customer engagements.

Participated · Series A · Jan 2025

SeeQC is building what it claims to be the first fully digital quantum computing platform, pairing classical and quantum circuitry on the same chip to improve speed, stability, and cost-effectiveness. Its architecture uses digital readout and control technologies implemented in a chip-scale, multi-layer superconducting design. Originating as a spin-out of superconductor electronics pioneer Hypres, the company also operates one of the earliest commercial multilayer superconductor foundries, giving it in-house design, fabrication, and test capabilities. By combining these assets, SeeQC aims to overcome the energy and scale limitations that hamper today’s analog quantum systems and to make quantum computing practically useful for global businesses. The firm plans to market turnkey quantum computing solutions and offer quantum-ready superconducting chips to partners. Although the company has not publicly disclosed revenue or user metrics, it recently secured a substantial $1.0 billion capital infusion, providing significant runway. The fresh funding will be directed toward accelerating R&D, expanding manufacturing capacity, and growing its team of quantum and superconductor experts.

Participated · Series C · Oct 2023

Seurat Technologies provides a 3D metal printing system designed for high-volume, decarbonized industrial production. The company intends to use new funding to accelerate its 3D metal printing technology and move from prototype to production. Seurat plans to deploy printing factories at or near customer sites worldwide to produce parts closer to demand, reshore supply chains, and reduce emissions from transportation and global logistics. Its pilot factory in the Greater Boston area will have capacity to produce more than 25 tons of metal parts annually and is already oversubscribed. Seurat has commitments to develop 59 tons of additively manufactured metal components for Siemens Energy over six years, LOIs from six customers totaling 4,000 tons of material, and over $750 million in projected revenue over the next several years. The company anticipates its technology could directly mitigate as much as 100 million tons of CO2 by 2030 and holds over 275 patents and trademarks, granted and pending. The company is led by CEO James DeMuth and is based in Wilmington, MA. Seurat Technologies provides scalable metal additive manufacturing technology and components manufacturing via its patented Area Printing platform. The Area Printing process decouples resolution and speed, enabling gains in lead time, cost, and quality. The technology originated at Lawrence Livermore National Laboratory; CEO and co-founder James DeMuth was a co-inventor and key developer. Seurat holds over 130 patents granted and pending. The process can be powered by 100% renewable energy and allows greater materials recycling, positioning the company to significantly reduce carbon emissions—0.15 GT of annual emissions displacement is expected by 2025. Seurat's commercialization program is slated to launch this year, and the company has received LOIs from leaders in automotive, aerospace, energy, and industrial manufacturing. Funding will support the initial commercial rollout, hiring, and development of a production-grade system. Seurat Technologies commercializes Area Printing, a metal additive manufacturing process that focuses over two million controllable laser points on a bed of metal powder to produce production-grade components. The process emphasizes per-point control of laser power and duration to deliver resolution, part integrity, and reliability at scale. Seurat says its first system generation will cut cost per part by about 50% versus today’s additive manufacturing technologies and that future printers could reach manufacturing costs below conventional die casting by 2030. The technology originated at Lawrence Livermore National Laboratory and was developed by a team that included Seurat’s CEO and co-founder James DeMuth. Seurat plans to use new funding to accelerate development and commercialization of Area Printing. The company is based in Wilmington, Mass., and previously raised $13.5 million in a Series A led by True Ventures in January 2018. Seurat Technologies builds industrial-scale metal printers aimed at cost-effective, high-throughput metal additive manufacturing. Its approach leverages technology originally developed at Lawrence Livermore National Laboratory and the company has applied for more than 20 patents. Seurat’s printers are designed for industrial scalability, likened to large-format laser engraving machines for manufacturing. The company intends to use new funds to accelerate commercialization of its metal additive manufacturing technology. Seurat was founded in July 2015 and is based in Boston, MA. Prior to this round it raised a $3.41M seed financing in June 2016 led by True Ventures.

Participated · Series C · May 2023

Lightmatter builds a purely optical interconnect layer that lets hundreds to thousands of GPUs work synchronously in specially designed racks. Its photonic chips and on-rack optical wiring can enable 1,024 GPUs to operate together and the current interconnect delivers about 30 terabits, with 100 terabit capability on the roadmap. The company has been developing photonic chips since 2018 and is also working on new chip substrates to enable tighter networking tasks using light. Lightmatter positions itself as a foundry-like platform for hyperscalers, and several major datacenter companies are already customers though unnamed. Management expects power-per-chip and wafer-scale approaches to be key future differentiators and is preparing technology roadmaps accordingly. The company’s tech targets the interconnect bottleneck that limits scaling beyond single racks and aims to dramatically reduce latency and power overhead in large clusters. Lightmatter develops photonic AI hardware that uses arrays of microscopic optical waveguides to perform matrix-vector computations with light, targeting large AI workloads. Its full-stack offering includes Envise (computing hardware), Passage (interconnect) and Idiom (software) that integrates with PyTorch and TensorFlow. The company licenses relevant patents from MIT and highlights the ability to create “virtual chips” by using multiple wavelengths to multiply computational density. Lightmatter says this analog-digital hybrid approach can substantially reduce energy consumption compared with traditional GPU/TPU architectures. The company is running several pilots in beta and plans mass production in 2024. Financially, Lightmatter has been scaling since an $11M seed in 2018 and additional capital in 2021, and most recently closed a $154M Series C. Lightmatter develops photonic computing hardware and systems — including the Envise photonic AI accelerator — that perform certain neural-network linear algebra operations by steering light through optical circuits rather than using transistors. Its chips are currently optimized for high-throughput linear algebra (inference) and the company says its platform can be ~5x faster than an Nvidia A100 on large transformer models while using roughly 15% of the energy. The firm has built server-sized units that fit into standard racks and is working on a wafer-scale optical interconnect called Passage for future generations to improve chip-to-chip communication. Lightmatter has offices and more than 70 employees in Mountain View and Boston and is preparing an early access program to deliver test units to hyperscale customers. The company is hiring to support hardware production, go-to-market, service/support and the software stack required to operate the systems. Lightmatter develops ultra-efficient light-powered AI chips built on photonic processor technology. Its optical computer chips leverage the properties of light rather than electrical signals to enable fast and efficient inference and training engines. The company intends to use the newly raised funds to continue development of its light-powered computing platform. Financially, it announced $22M in a Series A-1 round, bringing its total Series A to $33M. The round was led by GV with participation from Matrix Partners and Spark Capital; GV partner Tyson Clark joined Lightmatter’s board. The funding supports continued R&D to advance photonics-based inference and training capabilities. Lightmatter develops photonic chips that perform key AI matrix-vector operations by routing light through configurable optical elements to compute results at the speed of light with far lower power and latency than conventional silicon chips. The technology originated from research by CEO Nick Harris at MIT; MIT owns and licenses some of the related patents to the company. Lightmatter has demonstrated a prototype chip with 32 “neurons” and is working toward chips with hundreds of neurons to increase density and parallelism. The chips are intended for specialty AI hardware used by developers and cloud providers rather than consumer devices. The company says its approach is compatible with existing CMOS fabrication methods, avoiding the need for new fabs. The recent financing is intended to build the team that will take the technology from prototype to product and commercial deployment.