Radical Reads

Introducing the Radical AI Founders Masterclass Inaugural Compute Cohort

By David Katz, Partner

http://Screenshot%202025%2001%2026%20161212

Today we are excited to announce the first cohort of the Radical AI Founders Masterclass program selected to receive compute grants. 

The Radical AI Founders Masterclass aims to equip aspiring and emerging founders with the tools and expertise they need to transform groundbreaking research into world-class AI businesses. To better deliver on this goal, we announced a Radical AI Founders compute cluster in the fall of 2024. In partnership with Google Cloud, each selected applicant from the Radical AI Founders Masterclass community will receive access to cutting-edge infrastructure valued at over $250,000, which enables them to scale their AI models using  Google TPU and Nvidia GPU state-of-the-art chips. 

The large number of applications for this first cohort demonstrates the real need for compute resources at this earliest stage of commercializing AI research. Applicants were chosen from a pool of exceptional talent from top global institutions including Stanford, University of Toronto, Oxford and MIT, and leading AI organizations such as OpenAI, Meta, and Amazon. While only a small percentage of founders, representing a limited number of projects, were selected, the ambition of all applicants was truly inspiring. New and returning applicants are encouraged to continue refining their projects using the resources offered through the program, and are invited to apply to the next cohort of the program to be announced in the coming months. 

12 Radical AI Founders applicants were awarded compute grants, as part of this first cohort. While several of the recipients remain in stealth mode for now and will be shared at a later date, we are delighted to introduce a few of the startups that make up this inaugural cohort.

Ribbon

Founding Team: Dave Vu, Arsham Ghahramani

Location: Toronto, Canada

Co-founders Dave (ex-Local Logic, Ezra) and Arsham (ex-Applied Machine Learning Lead at Amazon) are building Ribbon to help large enterprises discover top talent faster with an AI interviewing platform capable of conducting voice-based screening interviews at scale. 

Vmax

Founding Team: Matthew Sargent, Augustine Mavor-Parker 

Location: London, UK / New York, US

Matthew and Augustine are AI researchers from University College London (UCL), where they earned their PhD in Machine Learning. They are pioneering the use of offline reinforcement learning and LLMs to uncover patterns between patient events and clinician actions in an effort to streamline clinical workflows. 

OnCallNinja

Location: Seattle, US

Drawing on extensive experience scaling cloud and ML infrastructure gained at major tech companies, this team is developing an AI-powered platform for automating incident resolution. OnCallNinja’s system integrates code, documents, logs, metrics, and chat data to generate root cause analysis reports, drastically reducing incident resolution time. 

FabricBio 

Founding Team: Cam Rzadki, Ian MacPherson 

Location: Toronto, Canada

Cam (MSc in Applied Computing from the University of Toronto) and Ian (PhD candidate in computer vision at York University) bring deep expertise in high-throughput screening and medical image processing. Their platform leverages AI agents to automate bioimage analysis for pharmaceutical and biotech companies to accelerate therapeutic discovery.

BRICS 4D AI
Founding Team: Srinath Sridhar

Location: Boston Area, US

Srinath Sridhar is an Assistant Professor of Computer Science at Brown University, Director of the Interactive 3D Vision Lab (IVL), and co-director of Brown Visual Computing. The Brown Interaction Capture System (BRICS) captures and processes 4D spatial data (3D + time) using specialized cameras, compute systems, and AI models to enable real-world spatial intelligence beyond traditional 2D approaches.

Vella.ai

Founding Team: Yonatan Feleke, Henok Ademtew

Location: San Francisco, US

Yonatan (ex-LinkedIn, Qualcomm) and Henok (a NeurIPS published ML Engineer) are building an AI-powered smart inbox that reimagines email through contextual smart cards for different use cases. Their platform leverages fine-tuned foundational models to make everyday internet experiences more convenient, starting with email.

FastChart AI

Location: Cambridge, UK

A team of researchers from the University of Cambridge  is seeking to revolutionize visual chart understanding with generative AI. 

 

We’re thrilled to support these startups on their journey to commercialize AI innovations. 

To learn more about the Radical AI Founders Masterclass program and the Radical AI Founders compute cohort, visit: https://radical.vc/masterclass/

AI News This Week

  • Xanadu claims a big step towards useful quantum computers  (The Logic)

    Radical portfolio company Xanadu has unveiled Aurora, the world’s first prototype of a universal photonic quantum computer. Published in Nature, the system connects 35 photonic chips across four server racks through 13 kilometres of fibre optics. With ambitions to build a quantum data center by 2029, Xanadu is already working with companies like Volkswagen to explore applications, including battery material simulation and cracking encryption standards. 

  • The second wave of AI coding is here  (MIT Technology Review)

    While first-generation AI coding tools could generate syntactically correct code, a new wave of systems aims to understand how developers think and work. These new systems analyze developer workflows and codebase context to understand programming logic and processes. Companies in this field, like Radical portfolio company Solver, are pioneering approaches that enable AI to drive projects from concept to completion. The technology is already reshaping how engineers work, with companies like Alphabet reporting that 25% of new code is now AI-generated.

  • A virtual cell is a ‘holy grail’ of science. It’s getting closer.  (The Atlantic)

    Researchers are leveraging large language model architectures to develop foundation models for cellular biology. These models aim to predict complex cellular behaviours and interactions by training on vast biological datasets. Early successes by companies like Radical portfolio company Nabla Bio demonstrate how generative drug design enables more targeted wet lab experimentation. While experts estimate full cell simulation is decades away, AI can extract biological patterns to accelerate hypothesis testing and experimental design.

  • Lessons from Canada’s Nobel Prize win, and why capping graduate students will harm our economy  (The Globe and Mail)

    University of Toronto President Meric Gertler argues that long-term investment in research and embracing international talent is crucial for innovation. Geoffrey Hinton, an AI pioneer who immigrated to Canada, helped establish Toronto as a global AI hub. The 2024 Nobel Prize Winner co-founded the Vector Institute alongside Radical Ventures co-founders Jordan Jacobs and Tomi Poutanen, and Hinton’s influence helped shape several AI companies, including Radical portfolio companies Cohere and Waabi. Gertler warns policies limiting international students could harm Canada’s innovation ecosystem.

  • Research: Universality of representation in biological and artificial neural networks  (MIT/Harvard/NYU)

    New research shows that artificial neural networks and human brains converge on similar information processing patterns despite their different architectures. By studying how AI models and humans process language and images, researchers found that stimuli creating similar responses across AI systems also trigger predictable brain responses. Their findings on how artificial and biological systems discover similar optimal solutions when solving complex tasks like visual processing or language understanding point to universal principles in how intelligence represents information.

Radical Reads is edited by Ebin Tomy (Analyst, Velocity Program, Radical Ventures).