Back

Archetype AI

Company Overview

Archetype AI is a Physical AI company pioneering a new form of artificial intelligence capable of perceiving, understanding, and reasoning about the physical world. Founded with the vision to encode the entire physical world, Archetype AI is building Newton, the first foundation model that can understand the physical world in real-time.

The company aims to solve real-world problems by revealing hidden patterns and behaviors that escape human perception using the power of AI. Archetype AI is based in the technology industry, focusing on information and internet applications of physical AI.

Products Overview

Archetype AI’s primary product is Newton, described as a Large Behavior Model (LBM). Newton is a first-of-its-kind AI model designed to understand the physical world. Key features of Newton include:

  1. Multimodal sensor data fusion: Newton integrates data from various sensors including accelerometers, gyroscopes, radars, cameras, microphones, thermometers, and other environmental sensors.

  2. Natural language integration: The model combines sensor data with natural language to provide insights about the physical world in real-time.

  3. Universal embedding space: Newton encodes sensor signals into a single compressed mathematical representation of the physical world, updated in real-time.

  4. Generative output: The model can decode its universal embedding space into various forms to help humans understand it, such as text, visualizations, or code.

  5. Natural interaction: Users can interact with Newton using natural language, speech, or gestures to get a stream of insights and predictions about the world around them.

  6. API access: Newton offers an API for developers to integrate physical AI capabilities into their applications.

  7. Sensor flexibility: The model can work with any combination of sensors and can process both real-time and pre-recorded data.

  8. Private data handling: Users can run Newton with proprietary data on Archetype’s infrastructure or their own, ensuring data privacy.

  9. Customization: Newton can be fine-tuned with proprietary data for specialized use cases.

  10. Edge deployment: Mini models can be generated from the large model for deployment on edge devices with specific compute and power constraints.

Founding Team

Archetype AI’s founding team consists of experienced professionals from leading tech companies and research institutions:

  1. Ivan Poupyrev, Ph.D. - CEO/CTO
    • 20+ years of experience in research and product development
    • Previously led technology development at Google’s ATAP, Walt Disney Imagineering, and Sony
    • M.S. in Applied Mathematics and Computer Science from Moscow Airspace University
    • Ph.D. in Computer Science from Hiroshima University
  2. Brandon Barbello - COO
    • 13+ years as a senior product leader
    • Previously managed privacy-preserving on-device AI projects at Google
    • Founded multiple companies in advanced sensing and fintech
  3. Leonardo Giusti, Ph.D. - Head of Design
    • 15+ years as an award-winning design and research director
    • Previous roles at Google’s Advanced Technology and Projects and Samsung Design
    • Postdoctoral Associate at MIT Design Lab
    • Ph.D. in human-machine interaction from the University of Florence
  4. Jaime Lien, Ph.D. - Head of Machine Perception
    • 12+ years developing and productizing novel perceptual sensing systems
    • Previously a Sensing Lead at Google’s Advanced Technology and Projects
    • Developed wireless communication and localization technologies at NASA’s Jet Propulsion Laboratory
    • B.S. and M.Eng. in Electrical Engineering from MIT
    • Ph.D. in Electrical Engineering from Stanford University
  5. Nick Gillian, Ph.D. - Head of AI & Software
    • 15+ years in innovation and productization of AI and interaction technology
    • Led machine learning research at Google’s Advanced Technology and Projects
    • Postdoctoral Associate and Fulbright Scholar at the MIT Media Lab
    • Ph.D. in machine learning for real-time gesture recognition

Problem and Market Fit

Archetype AI addresses the challenge of understanding and interpreting the complex physical world around us. The company argues that while current AI has been largely focused on digital problems, the biggest challenges in the world are physical. Traditional methods of understanding physical systems often rely on human perception and interpretation, which can be limited, biased, or unable to capture fast-moving or complex phenomena.

Archetype AI’s solution, Newton, aims to bridge this gap by providing a comprehensive AI model that can interpret data from various sensors, revealing patterns and behaviors that are typically imperceptible to humans. This approach has potential applications across multiple industries, including:

  1. Safety and security (e.g., monitoring construction sites)
  2. Smart homes and spatial computing
  3. Automotive industry (adapting vehicles to people)
  4. Manufacturing (predictive maintenance in factories)
  5. Energy sector (monitoring renewable energy infrastructure)
  6. Healthcare (generating comprehensive well-being reports from wearable sensors)

The company positions itself in the emerging “sensor economy,” which represents a significant market opportunity as more devices and environments become sensor-enabled.

Business Model

While specific details about Archetype AI’s business model are not explicitly stated, the company appears to be following a platform approach:

  1. Developer Platform: Archetype AI is inviting enterprises and developers to join their Design Partner program and start using Newton to solve real-world problems.

  2. API Services: The company offers API access to Newton, allowing businesses to integrate physical AI capabilities into their applications.

  3. Customization and Fine-tuning: Archetype AI provides services for customizing and fine-tuning Newton for specialized use cases, which could be a revenue stream.

  4. Infrastructure Services: The company offers options to run Newton on their infrastructure or the client’s infrastructure, suggesting a potential cloud-based service model.

  5. Edge AI Solutions: By providing the capability to generate mini models for edge deployment, Archetype AI could be targeting the edge computing market.

Funding and Runway

The provided material does not contain specific information about Archetype AI’s funding rounds or runway. Further research would be needed to determine the company’s financial status and investment history.

Competitive Landscape

The material does not provide explicit information about Archetype AI’s competitors. However, given the nature of their technology, potential competitors could include:

  1. Large tech companies working on AI and sensor fusion (e.g., Google, Apple, Microsoft)
  2. Specialized AI companies focusing on physical world applications
  3. IoT platform providers
  4. Edge AI companies

A more detailed competitive analysis would require additional research.

Customers

The material does not list specific customers. However, Archetype AI is targeting a broad range of potential clients across various industries, including:

  1. Construction companies
  2. Smart home technology providers
  3. Automotive manufacturers
  4. Factory and warehouse operators
  5. Renewable energy companies
  6. Healthcare providers and wearable device manufacturers

The company is currently inviting enterprises and developers to join their Design Partner program, suggesting they are in the early stages of customer acquisition and product validation.

Relevant News

  1. April 8, 2024: Archetype AI was featured in a WIRED article by Steven Levy, exploring how their AI models serve as a translation layer between humans and complex sensors.

  2. October 24, 2023: Archetype AI announced a partnership with Infineon. Infineon will be the first company to utilize the Newton AI developer platform, offering device makers a combined package of sensor hardware and sensor AI.

  3. April 5, 2024: Archetype AI officially introduced their company and vision for physical AI, unveiling their approach to understanding the real world in real-time.

These announcements indicate that Archetype AI is actively building partnerships, gaining media attention, and moving forward with their product development and market positioning strategies.

Classification: AI Tier 1

  1. Core AI: Create fundamental AI technologies/base models
  2. AI-Enabled: Core offerings rely on recent AI advances
  3. AI Adopters: Use AI to enhance existing products/services
  4. Non-AI: No AI in products/services

Archetype AI creates foundational AI technologies, exemplified by their Newton Large Behavior Model that integrates multimodal sensing with real-time physical world understanding.