Mar 2010 – Sep 2010 · 6 mos

• Search Products Group
• Working on infrastructure components for features in various search products including Multimedia search and Web search.
• Scalability expert, designing flexible and scalable low latency systems for handling requirements at Yahoo!'s scale.
• Mentoring team members, defining the engineering process, leading the development efforts.
• Mapping the product requirements to engineering requirements and defining the tasks for the team.
• Responsible for choice of technology, architecture and execution.

Oct 2008 – Mar 2010 · 1 yr 5 mos

• Multimedia Search
• Responsible for the entire Image Content Systems at Yahoo!
• Was leading the entire development and maintenance effort of image crawlers & feed processing systems
• Responsible for Thumbnail serving system with low latency for few billions of images for web search and multimedia search.
• Experience with apache web server, yahoo traffic server for real time serving systems

Apr 2007 – Sep 2008 · 1 yr 5 mos

• Yahoo! Image Search
• Designed and developed a new distributed system to crawl and store the images based on Hadoop to address the scalability concerns (billions of records)
• Experienced in deciding and deploying kind of hardware for hadoop, coming up with different tuning parameters for hadoop based on the types of the application and the scale.
• Incorporated many complex image processing algorithms such as face detection, adult content discovery into our content systems as mapreduce jobs
• Scalability, reliability, load balancing and redundancy expert.
• Expert in hardware choice, capacity planning for given requirements based on scale, qps and latency. Handled hundreds of machines for hadoop cluster setup and thumbnail server systems

Dec 2004 – Mar 2007 · 2 yrs 3 mos

• Yahoo! Desktop Search
• Member of team that developed Windows based desktop search product.
• Involved in the design and development of this product from the scratch.
• Worked on several modules mainly back end of the product in C++
• 1. A crawler to collect the data from the machine.
• 2. Module to keep the database in sync with the data collection of the user.
• 3. Information Retrieval – indexing and querying modules.
• 4. Relevance ranking of the results.
• 5. Multi Threading, configuration, io and utility and politeness modules.
• In addition to the ownership of the important modules I have also worked on the other modules for resolving the bugs and concurrency issues
• Was responsible for analyzing the reports from the field, troubleshooting and fixing issues.

Jul 2004 – Dec 2004 · 5 mos

• • Worked on a product called Bloomba -A very fast search-based email client. for StataLabs, later acquired by Yahoo!