Jude Nelson

Graduate Student
Department of Computer Science
Advisor: Dr. Larry Peterson

Snail Mail:
35 Olden Street
Princeton, NJ 08540-5233

Email: jcnelson@cs.princeton.edu

If you see this, it means the picture did not load

About Me

I defended my dissertation on May 22, 2018. Now I work at Blockstack PBC as an engineering partner.
Below is an overview of projects I worked on for my PhD.


Syndicate is a scalable software-defined storage system. It gives users the ability to manage data consistency guarantees, dataflow processing logic, and administrative and trust boundaries in an end-to-end fashion, across a dynamic set of existing services. By doing so, Syndicate can safely combine multiple existing cloud services into a coherent storage medium, and remain resilient to individual service outages, API changes, and cost changes. Applications using Syndicate do not need to evolve in lock-step with the cloud services they leverage, and application developers can mix and match services to find their cost/performance sweet spots.

Syndicate is the subject of NSF grant #1541318. The extended abstract can be found here. The whitepaper is here. The source code and wiki are here.


Blockstack is a jointly-developed system between myself, Muneeb Ali, and Ryan Shea '12. It is a scalable, decentralized, global database that binds human-meaningful names to data via an exisitng blockchain (currently Bitcoin's). It scales by storing data in existing cloud storage providers, while leveraging the blockchain to ensure end-to-end data authenticity, integrity, and consistency between readers and writers.

The whitepaper can be found here (USENIX ATC 2016), and the source repositories are here.


STEAK (Security-Transparent Email with Automatically-managed Keys) is an email service built on top of Syndicate. STEAK offers the usability semantics of Webmail with the security of PGP (that is, end-to-end confidentiality, integrity, and authenticity). It solves the key distribution problem in two ways. To distribute public keys, it replicates them to sufficiently many non-colluding services (i.e. Facebook, the Bitcoin blockchain, etc.) that an adversary can't change it undetectably. To distribute private keys, it seals each key with the user's password, and stores it in the user's Syndicate Volume (limiting access and hedging against ciphertext compromises). STEAK is pull-based, unlike SMTP: a sender hosts the messages in her Syndicate Volume, and makes them available to the receiver to download.

The whitepaper is here. The source code is here.


WISH is the Wide-area Interactive SHell. It extends familiar UNIX shell scripting semantics to manage wide-area hosts by augmenting an existing shell (like GNU bash) with an environment and command set to write and run parallel and distributed jobs. It provides commands for spawning, synchronizing, signalling, and controlling any process on any host in the network, as well as redirecting any process's input, output, and error to arbitrary hosts. It provides an environment for processes to get, set, and atomically test-and-set globally-visible shell variables, query the health of remote hosts, and expose local files for remote processes to read. All WISH-spawned processes print their output and error to the user's TTY and/or local files as if they were running asynchronously on localhost.

The whitepaper is here. The source code is here.

Classes Taken

  • Fall 2013
    • COS597G Surveillance and Countermeasures
  • Spring 2012
    • COS424 Interacting with Data
  • Fall 2011
    • COS518 Advanced Computing Systems
  • Spring 2011
    • COS423 Theory of Algorithms
  • Fall 2010
    • COS561 Advanced Computer Networks
    • COS402 Artificial Intelligence
    • COS441 Programming Languages
    • COS597B Advanced Topics in Datacenters

Classes TA'ed

Before Princeton

I received my BS in computer science from the University of Arizona in May 2010, with a minor in Mathematics.  I attended the University of Arizona on a President's Award for Excellence scholarship, awarded in 2006.  I graduated Summa Cum Laude with Honors.

My senior thesis was An Improved Multiprotocol Application Data Transfer Service.  I implemented an intelligent file transfer daemon (IFTD) and a protocol framework which allows it to use multiple unmodified file transfer protocol implementations to concurrently fetch pieces of files from one or more remote hosts. It measures and records pre-defined features (such as latency and bandwidth) of each protocol's performance, as well as file attributes (such as size, MIME type, etc.) of the data being transferred, to rank protocol utility for future file transfers with simple machine learning.  IFTD currently serves as the file transport framework for Raven, a provisioning service currently deployed on PlanetLab. My undergraduate thesis advisor was Dr. John H. Hartman.