Led · Seed · Jun 2026

Qubic develops Kinetic Inductance Traveling Wave Parametric Amplifiers (KI-TWPAs), ultra-low-noise, wide-bandwidth microwave amplifiers intended for reading out superconducting qubits and compatible with other qubit modalities. Its amplifiers derive nonlinear inductance from the transmission line material rather than Josephson junctions, aiming for a more robust, scalable device that operates near the quantum limit and supports multiplexed qubit readout. Qubic reports heat dissipation for its amplifiers is expected to be cut to under 0.1 mW per amplifier, addressing a key bottleneck in cryogenic quantum systems. The company has sold its devices to Quantum Machines and plans to scale manufacturing to meet demand from quantum computing and defense customers. Qubic is also developing an RF quantum sensing platform leveraging its amplifier technology for applications such as threat detection and covert communications, and intends to grow its team three-fold to achieve these goals.

Participated · Seed · Apr 2026

Glif is a creative AI platform built around a chat-based agent that consolidates the full range of generative media tools — image generation, video creation, music, lip-sync, voice transformation, and more — into a single conversational interface. Rather than requiring users to navigate multiple specialized tools, Glif's agent is trained with expert creative knowledge and can handle everything from AI headshots and logo design to cinematic action scenes, product videos, YouTube thumbnails, and viral video formats. It targets creators, marketers, and e-commerce brands looking to produce high-quality visual and multimedia content at scale, without needing deep technical expertise across individual AI models.

Participated · Series A · Apr 2025

Blumind is a Canadian deep‑tech AI start‑up based in Ottawa developing ultra‑low‑power all‑analog AI compute semiconductors, software and data‑driven system solutions. Its AMPL™ analog AI technology is designed to dramatically reduce power consumption and latency for always‑on AI applications; Blumind says its products consume up to 1000x lower power than available alternatives and use standard CMOS process technology. The company is commercializing a first‑generation AI processor for processing real‑time sensor data and is accelerating design of next‑generation analog AI chips for vision and large language model inferencing. Blumind plans to expand engineering and development teams, accelerate chip design and production, and support existing tier‑1 lead customers while establishing strategic partnerships. The company recently closed a C$20M Series A and is keeping the round open as it engages with strategic investors to further scale operations and commercialization efforts. CEO and co‑founder Niraj Mathur said the funding will help rapidly scale the business and bring efficient, low‑latency AI to the edge.

Participated · Series A · Aug 2024

Viggle AI is a Canadian startup that develops a 3D-video foundation model (JST-1) designed to generate controllable, physics-aware character motion for video. The company’s product lets users replace or animate characters by uploading source video and images, or by using text prompts, and is used by filmmakers, animators and game designers as a visualization tool. Viggle offers a free, limited model via Discord and its web app, plus a $9.99 subscription tier and a creator program for select users. The CEO says the model’s strength is controllability and more realistic motion versus pixel-based video generators; outputs are usable but still imperfect (shaky, expressionless faces). Viggle partners with cloud providers including Google Cloud for GPU/TPU access to train and run models, and the company says it is talking with film and game studios about licensing its technology. The startup recently raised funding to scale, accelerate product development and expand its team.

Participated · Equity · Sep 2023

Story is building an "IP blockchain" that lets creators assert ownership, set usage parameters, and enable permissioned licensing and remixing of content. The startup positions its platform as a way to capture value for IP owners amid generative AI and other uses of content. Story has run a free, closed beta and says more than 200 teams and “more than 20 million addressable IPs” are registered through partnerships with Ablo, Sekai and Magma. The company plans to use the new funding to continue building the product and is aiming for a commercial launch later this year. Financially, Story has raised a total of $143 million alongside its parent PIP Labs and is now valued at $2.25 billion post-money. The founding team includes CEO and co-founder S.Y. Lee and CPO Jason Zhao; Lee previously founded Radish, which sold to Kakao for $440 million. Story Protocol is building an open infrastructure layer and IP protocol to power collaborative creation and ownership of creative intellectual property on the internet. The platform provides provenance tracking, frictionless licensing, and revenue‑sharing primitives that let creators across prose, images, gaming, audio and other media track provenance and enable contributors to remix while capturing value. It aims to democratize IP creation via a global, extensible IP repository and to attract third‑party developers offering services such as crowdfunding, capital formation, IP discovery, licensing modules, AI content authentication and community growth. The project emphasizes a modular, open architecture and a decentralized design intended to remove platform risk for ecosystem builders. Story Protocol launched with advisory support from filmmaker David S. Goyer and has raised over $54 million in funding led by a16z crypto. Its founding team combines entrepreneurs and operators with backgrounds at Radish, Episode, Google DeepMind, Dapper Labs, Harmony Protocol and other consumer tech, AI and web3 ventures.