Hyperdyne is a startup whose mission is to create the first easy-to-use computational marketplace by leveraging its intellectual property: a programming model with a technical stack designed from the ground up to enable general purpose compute on a decentralized, heterogeneous, and dynamic network of edge-devices in a fault-tolerant, massively scalable fashion.
Hyperdyne develops the Hypernet protocol, which brings visionary decentralized applications to life using the unique and exclusive attributes of our API, including:
-The ability to analyze data securely, on-device. This means sensitive user data, like healthcare or financial data, can be analyzed while remaining private, secure, and in its original location.
-Massive over-the-internet computation. Hypernet enables world-class supercomputer performance on your own internet connected devices.
-The ability to analyze data as it is generated. This means data never has to be sent to the cloud or stored; it can be immediately analyzed and discarded.
-The programming model is robust to machines unpredictably entering and leaving the network. This means reduced technical limitations for buyers and sellers of compute power.
-An unlimited number of machines of any specification can be utilized in parallel. This means the network is infinitely scalable, and anyone with any machine can contribute to the network; the only constraint is bandwidth.
Our first round of funding was highly successful, and we will be able to comfortably continue to work on Hypernet for several years.
Our team is dynamic and interdisciplinary. The Hypernet protocol itself requires deep knowledge of several technical domains, and our team is reflective of the different modes of thought required to build the product. This makes for a highly stimulating and interesting work environment.
Operating in a $500b+ industry, our growth potential is huge, and the competitive landscape is relatively uncrowded.
Our engineering team is currently small but highly interpersonal and cooperative. Technical development is tracked in Jira. Milestones are currently set and evaluated on a weekly basis. Our processes will adapt over time as our team grows in numbers and experience.
The technical architecture encompasses peer-to-peer communication, smart contract development, as well as highly scalable and numerically stable approximation techniques. We are developing a modular technical stack that can be used to quickly build rich and highly interactive native applications designed to preserve data privacy via local data hosting.
Flexible, modular, and secure peer-to-peer file transfer infrastructure designed to serve as the backbone for a diverse application ecosystem. The goal of this project is to develop an open-source technical stack that can transfer encrypted and authenticated data files between peers in a fashion that is agnostic to the underlying network. For example, the protocol should be able to operate via centralized cloud server or through decentralized file-system hosted on a network of unreliable machines while exposing the same standardized set of API calls to client-side software hosted by peers wishing to utilize this functionality.
Smart contract development for the governance and curation of a large scale registry of computational resources. The project is well suited for individuals who are interested in real-world blockchain applications and have a talent for recognizing corner cases that can be exploited by bad actors.
Hyperdyne is building a new decentralized programming model and an associated, language agnostic, API designed to facilitate large-scale linear algebra operations over the internet on unreliable computational resources. This API is intended to serve as the Hadoop (or MPI) of applications that require data to remain on the device in which it was created and where the availability of the devices themselves is highly variable. The underlying mathematical principle on which the programming model is based on comes from the field of cooperative robotics. The programming model itself was invented by researchers from Stanford University who were interested in developing fault-tolerant algorithms for exascale computing.
We always strive to make decisions with a long-term, future-focused perspective. Therefore, we value thoughtful, detail-oriented, quantitative mindsets focused on producing results.
We have great new office space in the heart of Silicon Valley (Palo Alto).
We are developing novel computational methods at Hyperdyne and will periodically publish in peer-reviewed journals. Published work may be presented at conferences.
Interested in this company?
Skip straight to final-round interviews by applying through Triplebyte.