Data Scientist - ML Forecasting & LLMs

Kumar Mantha

Arlington, TX • 469-***-**** • ***************@*****.*** • linkedin.com/in/kumarmantha • github.com/cosmicbeing619 PROFESSIONAL SUMMARY

Data Scientist with 3+ years of experience building, deploying, and monitoring machine learning models including predictive models, forecasting pipelines, and feature- engineered production systems across banking and telecommunications industries. Reduced client financial penalties by 61% through a hybrid ensemble forecasting; won 1st place at UTA's Datathon with 82% predictive accuracy. Skilled in Python, SQL, machine learning, feature engineering, and statistical analysis and growing expertise in LLMs, RAG, and Agentic AI.

WORK EXPERIENCE

Senior Systems Engineer Data Analytics Infosys Limited Sep 2022 - Jul 2024

• Drove a 34% increase in user engagement by mining SQL-based usage data to understand user behavior patterns, translating findings into actionable feature recommendations for product teams.

• Eliminated 99% of daily data outages by engineering and continuously monitoring a contingency ETL pipeline with automated failover, reducing manual incident response time from hours to minutes.

• Processed and analyzed 1M+ daily records using PL/SQL and ServiceNow, enabling real-time visibility into operations across four enterprise applications.

• Built executive-facing Power BI dashboards and KPI reporting frameworks that consolidated revenue and operational metrics, reducing ad-hoc reporting requests by standardizing data delivery.

• Led quarterly cross-application migrations, collaborating closely with engineering teams and business stakeholders to ensure scalable integration and zero-downtime transitions.

Assistant Systems Engineer Data Analytics Tata Consultancy Services May 2021 - Sep 2022

• Ensured 99.9% uptime for NEFT and RTGS payment systems processing real-time transactions for the Reserve Bank of India, supporting millions of daily financial operations.

• Migrated 30M+ records daily using SQL-based disaster recovery pipelines, reducing recovery time and safeguarding data integrity across production systems.

• Improved IBM MQ message throughput by optimizing queue configurations and eliminating bottlenecks, reducing message processing latency.

• Automated recurring reconciliation and reporting processes using SQL and Shell scripting, saving an estimated 10+ hours of manual work per week.

• Mentored 7 junior engineers on SQL best practices, incident management workflows, and system monitoring tools, accelerating their onboarding by 4 weeks. PROJECTS & ACHIEVEMENTS

Scot Forge Capstone Project

• Evaluated and developed a hybrid ensemble LGBM + 95th percentile quantile regression model, exploring multiple algorithms and engineering calendar, lag, and rolling features from 10 years of historical data to maximize predictive performance.

• Deployed the forecasting model into a production workflow and reduced client financial costs by 42% by improving demand forecast accuracy, enabling Scot Forge to optimize production scheduling and avoid costly contractual overruns. 1st Place - UTA Datathon

• Won first place among competing university teams by building a LightGBM + Random Forest ensemble model achieving 82% predictive accuracy on a real-world business dataset.

• Deployed a user-facing application to present model outputs interactively to judges and stakeholders. JobHunter AI - Chrome Extension

• Built an AI-powered Chrome extension integrating LLM-based, agentic workflows to support human-AI collaboration in job discovery automating tailored cover letter generation and job-fit recommendations to proactively surface relevant opportunities. 5th Place - INFORMS Data Challenge

• Developed an end-to-end ML pipeline optimizing XGBoost, LightGBM, and Random Forest models, achieving 88% accuracy and placing 5th in a preliminary analytics competition.

EDUCATION

M.S. in Data Science - University of Texas at Arlington GPA: 3.9 / 4.0 Aug 2024 - May 2026 B.E. in Electronics - University of Mumbai GPA: 7.42 / 10 Aug 2016 - Oct 2020 TECHNICAL SKILLS

Languages: Python, SQL, PL/SQL, Unix Shell

Machine Learning & AI: Gradient Boosting models, Regression Analysis, Time-Series Forecasting, Feature Engineering, PCA, LLMs, Generative AI, RAG, Agentic AI Data Analytics: Statistical Analysis, Hypothesis Testing, A/B Testing, EDA, Data Modeling, Data Cleaning BI & Visualization: Power BI, Tableau, Advanced Excel, Marimo Cloud & Tools: Microsoft Azure, Apache Tomcat, IBM MQ, ServiceNow

Contact this candidate