Jan 2020 – Jan 2022 · 2 yrs · Bengaluru Area, India

• Worked in the Search Core Platform team. Designed & developed a Search query logging and debugging framework to help developers debug Search queries as a graph across multiple systems.
• Initiated and executed the strategy to break the monolithic Core Search service into microservices architecture to achieve better fault tolerance and reliability across all the Search queries.

Dec 2018 – Jan 2020 · 1 yr 1 mo · Bengaluru, Karnataka, India

• Built a ML based clustering solution to automatically resolve discrepancies during invoice processing. This helped to identify the issues in invoices such as incorrect amounts, double payments, etc. worth $10 million.
• Built a highly scalable fault-tolerant batch processing system using Apache Spark to calculate the shipping costs of all items delivered by Amazon. The system now does all the calculations in 50 minutes as compared to earlier time of 4 hours, thus reducing the processing time by 80%.
• Previously worked on designing and implementing a cell based architecture to process the shipping costs in a highly parallel and cost effective manner. The system identified over-scaled resources leading to a significant reduction in the fleet size and provisioned resources by 30%.

Jan 2018 – Dec 2018 · 11 mos

• Developed a real-time trade import application to enable traders to calculate risk metrics for a trade as soon as it is booked.
• The application reduced the time taken to calculate risk metrics for a trade from 4 hours to 3 minutes.

Jan 2017 – Dec 2017 · 11 mos

• Worked on fetching FX trades from front office desks, designing the trade flow and modeling the trade attributes for PnL calculation.
• Converted end of day mark-to-market feed to regional feed which reduced the latency in trade flow and sync by approximately 20 hours.

Aug 2016 – Dec 2016 · 4 mos

• Developed a web interface to query databases using plain English phrases instead of traditional SQL. Used Standford CoreNLP and SUTime libraries for natural language processing.

May 2016 – Aug 2016 · 3 mos

• The project was on high performance data cube materization and querying using Apache Spark. It consisted of querying mongodb database for medical imaging data, identifying dimensions of interest, generating data cubes, applying operations such as roll up, slice,etc. and finally storing it in data structures for future retrieval.
• GitHub Link: https://github.com/sameer9311/Data-Cube-Materialization

May 2015 – Jul 2015 · 2 mos · Mumbai Area, India

• Worked with the Enterprise Data Warehouse ( EDW - Core ) team to build a platform to provide single view of data accesses. Responsibilities included designing SQL queries, extracting metadata and restructuring the data sent to the presentation layer.

May 2014 – Jul 2014 · 2 mos · Mumbai Area, India

• Responsibilities included defining and applying evaluation metrics for the recommendation system and modifying the engine based on the evaluation results.