Vikrant Thakur Email & Phone Number

Gurgaon, Haryana, India

• Handled the onboarding of new accounts across card, lending and banking products by collaborating with stakeholders to add required intelligence for decisioning systems.• Owned third party bureau data, to enable variables across all markets for credit and fraud risk modelling and strategy use case.• Built and maintained globalised bureau specific variables with optimised logic across different markets (US and international) and portfolios (consumer/commercial) in a third party gateway for downstream consumption.• Created Utilities using Python to automate data validation and generate several statistical reports to facilitate analysis. Consistently leveraged tools like SAS and SQL for offline variable creation and validation. Analysed impact of adding new variables on decisioning model by building, scoring and evaluating different model versions using Gini comparison.

Gurugram, Haryana, India

Noida Area, India

• Orchestrated an entire flow for streaming data collection, transformation and storage using tools like Apache Kafka, Apache Nifi and Vespa.ai• Single-handedly designed various microservices to perform analysis and trigger events using Python and Nodejs.• Carried discussions with the client-side Design and Security teams for seamless integration of services.• Collaborated with the team to always deliver the application product within the timeline.• Improved application efficiency and search performance in Vespa by 10x times using multithreading optimisation.• Authored init scripts for remote servers to save hours of weekly time.• Researched and tweaked with an open-source tool, Apache Superset to add new features aligned with the project expectations.

In the initial phase, learned about stock market trading basics, developing trading strategies, performing scans and explorations, and backtesting in the Amibroker software.• Created trading strategies based on some of the market trend indicators and oscillators.• Incorporated trailing stop losses and target arrays in both the long and short positions, and optimized parameters in the strategies.• Performed pair trading and hedging, and developed some trading premises.• Used ML models like SVM classifier and other Regression models for enhancing strategies.

• Worked in the development and enhancement of functionalities of PRIME, a geoscience software.• Performed text mining on Well reports, used Regular Expressions to extract important data and developed a backtracking algorithm for searching and computing required parameters.• Used Machine Learning algorithms to interpolate grid points for plotting 3D surfaces, and for prediction of the Well parameters.• Improved the predictive power of ML models using hyperparameter tuning, and Data cleaning and imputation.• Integrated R-scripts in JVM, and improved process efficiency using multi-threading in Java.