Big Data Analytics & Data Science @ Purdue Pharma L.P.
Client Integration Manager, FX Trading IT @ UBS Investment Bank
Calcutta International School (CIS)
17 + Years of Industry Leading Experience in developing the vision, strategy and execution of cutting-edge analytics platforms for Large Scale Predictive Analytics & Data Mining. As the senior architect for Big Data Analytics, I was responsible of implementing the firm’s Award-Winning Analytics Platform supporting Market Research across Terabytes of proprietary data supporting a multi-billion-dollar enterprise research
17 + Years of Industry Leading Experience in developing the vision, strategy and execution of cutting-edge analytics platforms for Large Scale Predictive Analytics & Data Mining. As the senior architect for Big Data Analytics, I was responsible of implementing the firm’s Award-Winning Analytics Platform supporting Market Research across Terabytes of proprietary data supporting a multi-billion-dollar enterprise research needs.
The platform has received a wide recognition as for being a unique and innovative application of High Frequency Trading & Investment Banking Technologies to solving real-world Big-Data problems outside of Finance, delivering 100x+ Performance at 1/3 rd the cost and Extremely Low Maintenance.
Big Data Platform Awards & Recognition
2014: InfoWeek Most Innovative Companies in the US: Ranked 10: Analytics for Epidemiology Research
2015: InfoWeek Most Innovative Companies in the US: Ranked 22: Big Data Mining with High Frequency Trading Technology
2015: CT Quality Improvement Award: Silver Place for Purdue Pharma’s Analytics Platform
Areas of expertise:
- High Performance Computing
- Big Data Architecture and Data Mining
- Large Scale Machine Learning
- Algorithmic FX Trading
- Parallel Programming
Big Data Analytics @ Responsible for the conceptualization, design and implementation of the firm’s award-winning data analytics platform for Data Mining, Machine Learning & Predictive Analytics on Terabyte Scale Datasets in Real-Time within usually a millisecond-level Latency for 90% of Queries. I lead the IT Data Science group that has full ownership and responsibility for delivering predictive analytics and Big Data projects across functional & business teams within the organization.
The Big Data Project led to significant IT savings in excess of $ 4 million within a span of 1–1½ years and significantly increased productivity (queries that used to take 2 days + -- 1 week can now be completed within 10 seconds for 90% of cases)
Areas of proficiency:
- Statistics & Machine Learning: Strong background in Statistics & Machine Learning
- High Performance Machine Learning: Multi-Node, Multi-Core, GPU Based & Out-of-Core Machine Learning using R & Python and command line C/C++ interfaces
- Big Data & Hadoop: Working familiarity with the Hadoop Ecosystem (Hadoop, Hive, HBase, Pig) and Spark
- NoSQL Databases: Cassandra, Redis, KDB+ (In-Memory/Columnar), MongoDB
- Cloud Computing: AWS - (EC2, S3, EBS, Route53), Azure
- Cloud Based Analytics: AWS Redshift, Google BigQuery, Azure ML
- Visualisation Tools: Tableau, R (ggplot2), Python (eg., matplotlib), Unix (gnuplot), Delta Dashboard, Panopticon
Predictive Analytics Project Examples:
- Demand Forecasting: M&A Asset Forecasting and Monte Carlo simulations
- Marketing & Churn Optimization: Predict Physician prescribing patterns based on immediate and past characteristics
- Miscellaneous Learning Models: Optimize Physician Target Lists based on profiles of recent medical graduates
- Associative Rules Mining: Multivariate Analysis of Physicians to determine factors for Churn and Change in Rx
- Natural Language Processing: Classification of clinical trials & 8 million+ medical records using NLTK, ElasticSearch, etc From June 2013 to Present (2 years 5 months) United StatesAssociate Director, FX Trading & Connectivity @ Managed IT relationships for more than 125 FX Client Accounts (Institutional and Buy-Side Clients including major US and European Hedge Funds) for enabling them to trade on the UBS High Frequency FX Trading Platform. Deep knowledge of market fundamentals, spot and derivative trading instruments, front-to-back trade flow from trade-capture to settlement and a detailed knowledge of the trading platform internals (markups, tolerance, grace period, etc.)
Strong knowledge of the various technologies and Algorithmic Trading platforms used by top-tier Hedge Funds to implement high-frequency ultra low-latency trading architectures.
Trading Analytics & Reporting developed:
- Metrics / SMA Reports: Used extensively by the FX Algo Desk to analyze high level trends
- Rejects Reports: Led to significant increase in trading volumes due to analysis and reduction of reject ratios
- Alerts System developed in KDB+ and R using ggplot2 enabled Sales Desk to receive early feedback on trend changes
Market & Technical Knowledge:
- Broad understanding of the Global FX market, ECNs, Algorithmic / High-Frequency Trading, Low Latency networks, Market Data feeds, Smart Order Routing, Data Analytics (SQL, KDB), FIX messaging, Standard and Specialised (FPGA) Feed Handlers From April 2008 to May 2013 (5 years 2 months) Associate Director, Investment Solutions Trading @ Design of business logic for discretionary and non-discretionary portfolio trading products, mainly in the area of trading hedge funds and other AI instruments.
- Managed Portfolio of Mutual Funds (MPMF) – Design of the Portfolio Execution Tool functionality
- NDT (Non Daily-Traded) Instruments in MPMF – Design of Hedge Fund of Funds Rebalancing
- Event-Driven Rebalancing – Design of rebalancing logic based on corporate “events” From July 2007 to April 2008 (10 months) Client Integration Manager, FX Trading IT @ Integration of client trading platforms with the UBS FX Trading platform using middleware, viz., IBM Websphere MQ, Tibco EMS and Tibco RV
Product Owner for FX Connectivity Center – A product-neutral STP solution for Trade Confirmation (eg., FXPB Give-Ups), Routing & Trade Enrichment for trades done via UBS DPP, OneTool and other FX & Fixed Income e-trading platforms From March 2005 to July 2007 (2 years 5 months) ECommerce IT Specialist @ Worked on more than 150 client projects in over 4+ years at IBM deploying Java Based Web Services solutions using Websphere Application Server and integrating them with backend systems such as DB/2, Lotus Domino, MQSeries, Ariba and other B2B/B2C components running on Apache-based IBM HTTP Server (IHS). Technical owner for IBM Support Site (http://www.ibm.com/support) which catalogued the IBM knowledgebase using advanced data mining tools. From May 2000 to November 2004 (4 years 7 months) Unix Capacity Planning Analyst @ AIX Administrator for the Philip Morris Management Corp (PMMC) and Philip Morris International (PMI) production environment. From April 1998 to May 1999 (1 year 2 months)
Nataraj Dasgupta is skilled in: R, KDB+, jQuery, Perl, Big Data, Python, Hadoop, Statistics, Data Mining, AIX, Java, Unix, Equities, FX Options, Integration, Machine Learning, Trading, Hedge Funds, Low Latency, Electronic Trading, Trading Systems, FX trading, Risk Management, Market Data, Trading System, Portfolio Management, FX Trading
Looking for a different
Get an email address for anyone on LinkedIn with the ContactOut Chrome extension