SUMMARY:

PhD Applied Mathematician with extensive expertise in trading strategies and software development, statistical analysis, quantitative analysis, financial and mathematical modeling, and more.

TECHNICAL SKILLS:

Applied Mathematics: Mathematical modeling using differential equations, finite differences, finite elements, asymptotic analysis, statistical analysis, optimization theory and more.

Machine Learning Techniques: Supervised learning: linear regression, k - nearest neighbors algorithm, Na ve Bayes Algorithm, decision trees, logistic regression, neural networks and support machine vectors, neighborhood component analysis Unsupervised learning: k-means algorithm, principal component analysis and Confidential

Big Data: Parallel computing, MapReduce (Hadoop or any database), data-storehouse

Financial Engineering: Development of statistical arbitrage based strategies for all equity types.

Software Development: Design and development of software packages/models for companies like Abbot cardio- vascular, Schlumberger, Saudi Aramco, Exxon Mobil, and Sandia National Laboratories.

Operating Systems: UNIX, Linux, MAC OS, DOS/Windows.

Programming Languages: MATLAB, Mathematica, Excel, JAVA, MySQL, Weka, R, SciPy, NumPy.

Platforms: Rediplus, Trader Work Station, Bloomberg, AWS, EC2, S3, Hadoop ecosystem, Hortonworks.

PROFESSIONAL EXPERIENCE:

Confidential, Austin, TX

Manager

Responsibilities:

• Quantitative investment management.

Confidential, Austin, TX

President

Responsibilities:

• Development and implementation of statistical arbitrage strategies.
• Analyzing large data sets involving different data types. Typical data set contains a minimum of 100K rows but could be millions of rows.
• Developing algorithms to mathematically test different competing ideas to select the best solutions based on predefined KPI’s using historical data sets.
• Using machine learning and data mining techniques to select the optimal model parameters based on historical data sets.
• Developing automated trading algorithms that takes as input continuous real time data flows, analyses the flows to determine the optimal real-time parameters for trading model.
• Validating all models and algorithms to ensure conformance.
• Deploying automated algorithms and models across board using relevant tools.
• Developing JAVA wrappers around heritage codes to run on recent platforms.
• Developing predictive models for consumer relationship management.

Confidential

Responsibilities:

• Lead a team in developing an Confidential system for the carbon dioxide removal system of the then largest ammonia plant in the world.
• The process involved using principal component analysis to reduce the variable space significantly while still capturing over 90% of the variance in the system.
• The reduced variable space was then used to design a system with the capacity to flag abnormal working conditions.

Confidential

Responsibilities:

• Trained a SVM algorithm to flag spam emails using data from spamasssing.apache.org.
• The performance of the model on the training set was in excess of 99% with the validation set giving an accuracy of 98%.
• Currently in the process of training a logistic regression model and neural network to perform the same classification.
• Trained a neural network to read handwritten numbers with an accuracy of over 98%. Model trained with digitized data set of pictures.
• Developed a Naïve Bayes mobile phone spam classifier with an accuracy of 97.6%.

Confidential

Responsibilities:

• Currently developing a recommender system to propose “selections” to different users based on previous use of certain products using regularized linear regression.

Confidential

Responsibilities:

• Developed a MapReduce scheme to process large text files, about 3GB in size. A typical file is highly unstructured containing names addresses, dollar amounts, etc.
• The code completes the assigned task in less than a minute on a cluster of 4 nodes. The system developed can be used on the
• Hadoop distributed file system or any other data repository.
• Although the typical file size is currently 3GB, the system can be scaled up to handle bigger files.

Confidential

Responsibilities:

• Designed an online trading platform for financial instruments.
• The model parameters are “learned” continuously in real time using “tick data” so that current market conditions are always reflected in the model decisions.

Confidential

Responsibilities:

• Modeled the structure - activity relationship of trimethoprim analogues binding to dyhydrofolate reductase utilizing an SVM algorithm.

Confidential, Evansville, IN

ASEE/NSF Industrial Postdoctoral Fellow

Responsibilities:

• Polycarbonate co-polymer purification/isolation process development and improvement using experimental and numerical techniques.

Confidential

Advanced Process Control Engineer

Responsibilities:

• Advanced process control on the largest ammonia plant in the world.

Confidential, Austin, TX

Financial Engineer

Responsibilities:

• Robust portfolio optimization.

Confidential, Austin, TX

Post Doctoral Fellow

Responsibilities:

• Developed a model for studying sand production in unconsolidated reservoirs.

Confidential, Albuquerque, NM

Research Intern

Responsibilities:

• Developed an algorithm to numerically measure the rotational and translational diffusivities of spheroidal particles in a solvent in C++.
• Implemented algorithm for both single and parallel processing.

Confidential

Responsibilities:

• Designed molecular dynamics simulations that would typically run on multiple computers over many days to study the behavior particulate suspension flows.
• The typical cluster size ranged from 24 to 48 nodes. Parallel computing was utilized in this project because the problem was embarrassingly parallel - little or no interaction between the “bins” utilized to parameterize the system.

Confidential, Houston, TX

Production Engineering Intern

Responsibilities:

• Developed integrated production models for the Exxon-NNPC JVC fields.
• Performed extensive reservoir fluid analysis.
• Developed and validated a dynamic reservoir model.

Confidential

Research and Teaching Assistant

Responsibilities:

• Algorithm development in MATLAB for thermodynamic and process control applications.