Siddharth Somani Email & Phone Number

Ahmedabad, Gj, In

Mumbai, Maharashtra, In

Developed an A/B testing model using Bayesian modeling and MCMC sampling for evaluating new strategies in Adidas' digital marketing initiatives.Enhanced model performance through advanced statistical techniques like parameter tuning and winsorization.Validated model accuracy with rigorous A/A testing, achieving a 4% error rate across 150 test cases, ensuring relaibility

Boston, Ma, Us

I was a team member in the Quantitative Development Team for Asset Management. I contributed to building high quality, robust, and efficient analytical solutions that will be used to improve the company's investment processes.Developed Options Sentiment Dashboard in R-Shiny and associated transformations in R.Enhanced time series models for Sovereign Debt Model representing 10 years of data, created visualization of patterns on R Shiny UI.Automated R-Shiny applications deployment by creating CI/CD pipelines using Jenkins and enabling faster deployment of UI updates.Streamlined automation for Data Acquisition, Data Abstraction Layer and Transformations by migrating the code to GitHub, developing CI/CD pipeline, and scheduling batch jobs using Autosys.

Worked towards finding an equation describing the relationship between metrics and log analytics table in Microsoft Azure ADLS Gen2.Collaborated with the data analytics team to create dashboards and email alerts for storage in Microsoft Azure. These dashboards will enable the team in forecasting storage leading to cost optimization and improving of user experience.

Minneapolis, Mn, Us

Created Path Comparison algorithm using RFID data of stolen items, to enable National Investigation Centre to speed up weak spot identification process 5x times.Developed backend by writing APIs using Java Spring Boot to query the database, streaming with Kafka and created test cases.Developed Front-end website using ReactJs to display aggregated information like map of travelled path, time, amount, etc. of more than 10000+ items stolen from 1800+ stores each month.Championed migration of CI/CD services from Drone to Vela. Also created standard methods to simplify migration efforts of a 10-member team.Automated Kubernetes commands using Python to report and correct the daily errors in pods in 1800+ stores.Resolved incidents raised by Stores team for Stolen Item Reporting Application.

Chicago, Il, Us

Improved the performance and logic of stored procedures written in SQL for SSRS reports.Created SSIS packages to check the correctness of data coming through ETL.Monitored file transfer using code written within R.Improved system security, by analysing and correcting the hardcoded word password.Maintained corporate website hosted on AWS.Worked using Agile methodology and participated in its various activities like daily sprints, backlog refinement, etc

Ahmedabad, Gujarat, In

Determined the effect of 8 different macroeconomic variables and their time lag on the returns of Indian Equity Markets using linear regression over data of past 10 years – Identified variables with positive, negative and no impact.Relevant data of the stock returns and the different factors assumed to affect the returns was collected from Bloomberg and other sources. Data was cleaned using statistical and numerical methods to make it ready for analysis. Cleaned data was analysed using regression analysis in R to find a relationship among the variables and the total returns, and for finding the risk factor. Using the derived equation conclusions were made for each variable on how it affected the stock market returns.

London, England, Gb

Identified approaches to be used towards finding solutions in the area of predictive maintenance of rotating equipment by using concepts of statistics and machine learning (neural networks).Learnt about basic concepts of statistics, signal processing and machine learning and its use in data analytics.Statistics: Stationary and non-stationary series, random walk, p test, t test, f test, central limit theorem, normal, chi squared and t distribution, 1-way and 2-way regression on ANOVA, correlation.Signal processing: Moving average and exponential filter, low pass and high pass filter, dft, fft, nyquist theorem, z transform and chirp z transform.Machine Learning: Neural NetworksCollaborated with the Energy Intelligence Team to determine if the machine was faulty by analyzing identified machine data.

Master Of Science - Ms, Applied Mathematics And Statistics | Quantitative Finance

Bachelor'S Degree, M.Sc. Mathematics (Dual Degree)

Bachelor'S Degree, B.E. Electronics And Instrumentation (Dual Degree)