Yashasvi Kumar

Data Analyst

Faridabad, Delhi NCR 0+ yrs exp 87 · Excellent

About

Engineering graduate (Thapar Institute, 2025) with a Dual Certification in Data Science & AI and hands-on analytics experience across consulting, banking and data science. At Ernst & Young, built a Python ETL and validation engine with a KPI dashboard that cut manual review effort by ~60%. At First Quadrant Labs, delivered two end-to-end analytical projects on datasets of 225,000+ and 230,000+ records — including identifying data leakage that would have produced a misleading model. Strong in SQL, Python, Advanced Excel and Power BI, with banking domain knowledge from HDFC Bank (FLEXCUBE, KYC/AML). Focused on validating data before analysing it, and translating findings into decisions non-technical stakeholders can act on.

Skills & Expertise (41)

SQL Intermediate

6.2/10

Years Exp

Data Cleaning Intermediate

6.0/10

Years Exp

Data Validation Intermediate

6.0/10

Years Exp

ETL Pipelines Intermediate

6.0/10

Years Exp

Python Intermediate

6.0/10

Years Exp

NumPy Intermediate

6.0/10

Years Exp

Pandas Intermediate

6.0/10

Years Exp

Data Modelling Intermediate

5.8/10

Years Exp

Statistical Analysis Intermediate

5.5/10

Years Exp

Exploratory data analysis Intermediate

5.5/10

Years Exp

scikit-learn Intermediate

5.5/10

Years Exp

Plotly Intermediate

5.5/10

Years Exp

Seaborn Intermediate

5.5/10

Years Exp

Matplotlib Intermediate

5.5/10

Years Exp

Classification Intermediate

5.5/10

Years Exp

Regression Intermediate

5.5/10

Years Exp

Model Evaluation Intermediate

5.5/10

Years Exp

Tableau Intermediate

5.5/10

Years Exp

Power BI Intermediate

5.5/10

Years Exp

NLP Intermediate

5.5/10

Years Exp

Requirements Gathering Intermediate

5.0/10

Years Exp

Gradio Intermediate

5.0/10

Years Exp

AML Compliance Intermediate

5.0/10

Years Exp

stakeholder communication Intermediate

5.0/10

Years Exp

Technical Documentation Intermediate

5.0/10

Years Exp

Git Intermediate

5.0/10

Years Exp

GitHub Intermediate

5.0/10

Years Exp

Hugging Face Intermediate

5.0/10

Years Exp

KYC Intermediate

5.0/10

Years Exp

Banking Operations Intermediate

5.0/10

Years Exp

MIS Reporting Intermediate

5.0/10

Years Exp

KPI Dashboards Intermediate

5.0/10

Years Exp

Advanced Excel Intermediate

5.0/10

Years Exp

Hypothesis Testing PowerPoint Streamlit MongoDb MySql PEFT LORA SQL Server

Work Experience

Data Science Intern

First Quadrant Labs

Jul 2026 - Sep 2026

Network Traffic Classification: built an ensemble intrusion-detection pipeline on 225,000+ network flows (CICIDS2017, 78 features) — resolved infinite flow-rate values, removed 2,633 duplicate records and 10 zero-variance columns, and benchmarked three classifiers to 99.99% accuracy and F1 on a stratified held-out set, reducing false positives. Accident Severity Prediction: developed a Gradient Boosting model on 230,000+ real-world records with 108 engineered features, achieving ROC-AUC 0.84 (5-fold CV 0.841 ± 0.009) on a severely imbalanced dataset (0.44% positive class). Identified and excluded outcome-leakage fields that would have inflated the score to 0.92 while making the model unusable in deployment; recommended an operating threshold capturing 79% of fatal cases within 31% of flagged records, delivered through a stakeholder report, executive presentation and interactive dashboard.

Assistant Manager, Retail Branch Operations

HDFC Bank

Nov 2025 - Mar 2026

Completed HDFC Bank’s ACE Banker Program (NIIT) with certification on FLEXCUBE core banking software, covering Liability & Asset Products, Compliance (KYC/AML), Capital Markets and Digital Banking. Processed customer transactions on FLEXCUBE to a zero-defect standard, maintaining KYC-compliant, audit-ready records; reconciled daily and investigated every discrepancy to source before close, escalating with full documentation.

Artificial Intelligence Intern, Tax Technology & Transformation

Ernst & Young (EY)

Dec 2024 - Apr 2025

Built a Python engine (pandas, NumPy, NLP) to extract, validate and reconcile data across PDF, Excel and CSV sources, flagging mismatches automatically — reducing manual review effort by ~60% and improving reporting accuracy. Designed automated ETL pipelines and an interactive KPI dashboard, converting raw multi-source data into decision-ready metrics for client and internal stakeholders. Gathered requirements with cross-functional stakeholders, supported testing and deployment, and authored technical documentation enabling handover and continued use.