JASPREET SURI

Currently leading a team of 24 software engineers responsible for delivering technically resilient and efficient (high precision/ high recall) surveillance models in hadoop map reduce. Responsible for managing 4 key surveillance portfolios (Market Surveillance, Know Your Customer, Global Investment Research and Insider Threat) and 35 distinct surveillance products within Financial Crime Compliance Engineering.

Technical responsibilities

• Code architecture finalization/ sign-offs
• Code reviews (selected as one of the code reviewers in a code review boot camp run by Compliance Engineering)
• Modeling and feature selection to improve precision/ recall

People management

• Direct manager for 4 Vice Presidents
• Local oversight of Financial Crime Compliance Engineering in Bangalore; part of Compliance Bangalore leadership.

Stakeholder management

• Understanding the surveillance requirements and discussing feasibility of technical solutions.
• Expectation management around release timelines, project prioritizations.
• Project planning; execution and post release support.

Responsible for delivering 4 Compliance products from scratch by understanding the regulatory risk and developing the technology product in hadoop map reduce. Promoted to Vice President/ Technical Lead (3 promotions) within 3.5 years of joining the firm as analyst.

Entity Resolution Surveillance:

• Replaced a vendor product responsible for performing name matching screening for client onboarding.
• Optimal technical architecture to reduce run time of entity resolution model across two large datasets (brute force approach would have resulted in 1mn* 1mn comparisons). The technology product was able to run everyday across the entire dataset which was a limitation in the vendor model.
• Model performance exceeded street standards (reduced volume by 100x while keeping recall above 95%)
• Currently working on real time implementation of entity resolution surveillance.

SCORE:

• Risk ranking tool built for Compliance to access risk profile for new and existing clients.
• Designed real-time risk ranking calculation infrastructure using RabbitMQ based event processing in combination with data storage on elastic cluster.

Unusual Movement of Assets Surveillance:

• Implemented transaction monitoring surveillance to process large transactions data and detect potential cases of money laundering in hadoop map reduce.
• Designed model features in collaboration with Compliance and trained multiple machine learning models on tagged data to optimize for precision and recall.

Employee Behavior Deviation Surveillance:

• Designed and implemented first of its kind behavior deviation model to detect unusual behavior patterns in employees which can serve as a predictor for insider threat risk.
• Feature design and data analysis in R.

• Worked on optimizing power consumption for existing and next generation WiFi chipsets.
• Awarded 2 Qualstars and Exceeds Expectation rating (highest among all campus hires) for independently supporting product lines for LG/ Samsung/ Huwawei.
• Received the opportunity to work for 3 months at the Qualcomm headquarters in San Diego within 6 months of joining the firm.