I currently work as a Senior Softwaree Engineer at Dropbox, New York with the Previews Infrastructure Services Team. Previews-Infra is the second largest fleet in dropbox, and provides middle layer services to convert uploaded files into previewable content for various services. Previously, I used to work at NEC Labs America, Princeton, NJ with the Systems Research Group (formerly a part of the Autonomic Computing Group). My research interests span data analytics, big data, stream processing, distributed systems, large scale system debugging, and program analysis. I have briefly also worked on cloud computing and software defined networking. I have a PhD in Computer Science from Columbia University, where I worked at the Programming Systems Laboratory with Prof. Gail Kaiser.
Professionally, I have also worked at the Systems Analysis and Verification Dept, from Jan 2010, to Sept 2010, and then from Jun-Aug 2011 at Princeton, New Jersey. I have briefly worked as Business Analyst at McKinsey & Co., New York in 2008. In my undergrad years, I interned as a Research Consultant at Instituto de Soldedura Equalidade (Lisbon, Portugal), a research organization under the aegis of the European Union where I was involved in a Project called “Natrualhy”. I was also a Research Assistant at the Indian Institute of Technology (Delhi, India) in the Computer Integrated Manufacturing Lab, where I worked on Supply Chain Management.
Pending patents available on request.
Most modern day softwares generate human readable logs for developers/administrators to understand and realize the cause of any error or behavior of the system. However, both the volume, velocity and non-uniform log formats make it difficult for administrators to easily find root-cause of errors in their systems. NGLA is a log analytics framework which automatically detects log patterns and leverages these patterns to give state-of-the-art automated real-time log anomaly detection
Modern computer systems, from single servers to large cloud deployments, generate billions of events that reflect the state and operation of the system. CLUE provides a black-box, unsupervised debugging tool to mine event patterns and diagnose performance issues in these systems. CLUE uses novel data mining technologies for automated information retrieval and a state-of-the-art debugging toolset to integrate and profile event transactions.