Daniel's projects

Home » Projects

I am interested in building secure, distributed and scalable systems.

Relevant experience:

• Systems security:
  • Final year project with Dr. Robert Watson on compartmentalisation of cryptographic components
  • Compartmentalised RSA functions of OpenSSL using the OpenSSL engine abstraction
  • Won 'Highly commended dissertation' award
• Scalable distributed systems:
  • Dropbox (2015) - Vacuuming team intern
    • Rewrote petabyte-scale garbage collection system for deleted Dropbox files in Go.
    • Implemented a distributed jobs service running across hundreds of dedicated machines, with a robust and efficient streaming pipelined design.
  • Facebook (2014) - Core Data team intern
  • Improved cache consistency in widely distributed storage systems (Memcache backed by MySQL DBs, thousands of machines spread across multiple continents).
  • Resolved race condition in cache invalidation mechanism that caused read-after-write inconsistencies in geographical regions with multiple MySQL database replicas.
• Deep learning research:
  • Leveraging deep reinforcement learning, policy gradients and LSTM networks to automate deep neural network design and optimize hyperparameters.
  • Building a virtual data scientist and experiments platform extending ModStore, a cloud-based machine learning platform which I previously built a scheduler for.
• Graph algorithms research:
  • Published clustering algorithm research with Dr. Li Xiao-Li and Dr. Ng See-Kiong at the Institute of Infocomm Research, A*STAR
• Machine learning:
  • Hands-on experience with deep learning, recommendation systems, supervised learning and unsupervised learning (clustering) algorithms.
  • Competitions: ACM SIGKDD Cup, Quora ML Codesprint, The Hut Challenge (recommendation systems challenge by The Hut Group)
• CTFs: MIT CSAIL-Cambridge C2C 2016 - 2^nd, Facebook-Cambridge CTF 2015 - 1^st
• International Olympiad in Informatics medallist

More details on: LinkedIn