Andrei Deftu

Main contributions on continuous integration and deployment methodologies. I established a standard development workflow (gitflow) with emphasis on automation and auditability. I applied this workflow into practice by designing and implementing an automated staged deployment framework to rapidly create effective business applications from GitHub, to TeamCity, to Artifactory. The applications were built and deployed as Docker container sets together with environment specific configuration via Puppet.

Measuring and optimizing transportation costs for real-time online orders. Achievements: architectural design, database setup and transactions optimization, retention and partition roll-over policies, implementation of a low-latency, high-throughput service to handle real-time demand, code review, interns mentoring, designed, implemented and managed a sub-project for deep diving into all stages the placement algorithm that was used by both the development and business analysis teams.

Prototyping and developing a scalable framework for testing various network protocols and services. Building a robust scheduler for running checks ensuring fairness and good throughput (C++, Python, Thrift, Boost, CMake). Prototyping and implementing an IP events tracker (Scala, Redis). Prototyping and implementing a scalable algorithm using CRDTs (Conflict-free Replicated Data Types) to ensure eventual consistency across replicated clusters (Java, Scala, Redis), published at ICDCS, IEEE, Jul. 2013.

I prototyped, implemented, and evaluated a resource pricing scheme and a highly-scalable allocation algorithm in game theory markets for large distributed systems on PlanetLab. Main challenges and achievements include: implementing a scalable, fault-tolerant, dynamic, and distributed algorithm for large networks, formal proof for distributed deadlock prevention, complex analysis of the mathematical model. Technologies: Java, Python, DHT.

Operating Systems Utilization Course. Labs on Linux, shell scripting, network administration and configuration. Operating Systems Course. Labs on fundamental OS concepts and system programming.