Software Engineer, AIML Engineer, Data Engineer, Data Scientist

Vinay Hulsurkar

+91-942******* # **************@*****.*** ï linkedin.com/in/vinayhulsurkar

§ github.com/KaizenVH24 Ð leetcode.com/u/vinayhulsurkar24 EDUCATION

Dr. D. Y. Patil College Of Engineering And Innovation Pune, India Bachelor of Engineering (B.E.) - Computer Engineering, Honours in Data Science 11/2022 - 06/2026

– CGPA: 8.66/10.00

– Awarded a full academic scholarship for the 4-year engineering program.

– Ranked Top 10 across the collge in second year university examinations.

– Secured Rank 1 in college for Database Management Systems (DBMS). TECHNICAL SKILLS

Programming Languages & Databases: Python, SQL (MySQL, PostgreSQL), JavaScript, TypeScript, HTML5, CSS3, MongoDB, SQLite

Machine Learning & Artificial Intelligence: Deep Learning, Predictive Analytics, Natural Language Processing (NLP), Retrieval-Augmented Generation (RAG), Large Language Models (LLMs), Time-Series Forecasting, K-Means Clustering, Model Context Protocol (MCP)

Frameworks, Tools & Libraries: Scikit-learn, TensorFlow, PyTorch, Keras, Pandas, NumPy, Ollama, LangChain, Hugging Face, ReactJS, NodeJS, Pydantic, VS Code, Jupyter Notebook, Google Colab MLOps, Cloud Infrastructure & Deployment: MLOps, CI/CD pipelines, Docker, Kubernetes, Git, GitHub, Linux, AWS Services (Conceptual understanding of SageMaker, EC2, S3), RESTful APIs, FastAPI, Flask Data Engineering & Business Intelligence: Exploratory Data Analysis (EDA), Data Preprocessing, ETL Pipelines, Streamlit, Plotly, Tableau, Power BI

WORK EXPERIENCE

Mainflow Services & Technology Pvt Ltd Remote

Data Science Intern 01/2025 - 02/2025

– Architected and deployed a K-Means clustering algorithm in Python to segment over 100,000 consumer behavioral profiles into 5 distinct cohorts, enabling highly targeted data-driven marketing strategies that increased projected campaign engagement by 22%.

– Designed and validated a highly performant Multiple Linear Regression model for real estate valuation, conducting rigorous A/B testing and hyperparameter tuning to minimize Root Mean Square Error (RMSE) and improve predictive accuracy by 18% over legacy baseline metrics.

– Led end-to-end data processing initiatives, utilizing Pandas and SQL to execute robust ETL (Extract, Transform, Load) pipelines, cleansing and normalizing 100,000+ transactional records for advanced exploratory data analysis

(EDA).

PROJECTS

Personal Finance Intelligent System - Python, Scikit-learn, Streamlit, Plotly Live App

– Deployed an Isolation Forest machine learning model for automated anomaly detection across raw transactional datasets, identifying fraudulent expenditure patterns with 94% precision and reducing manual financial review time by 40%.

– Developed interactive, real-time data visualization dashboards utilizing Streamlit and Plotly, enabling stakeholders to monitor dynamic Financial Health Scores and synthesize complex analytical data into actionable insights. House Worth & Credit Analysis - Python, Pandas, Scikit-learn, Streamlit, Plotly Live App

– Trained and optimized a Random Forest regressor to predict property valuations across 10 regional markets, achieving a highly accurate R-squared value of 0.97 and outperforming standard baseline metrics.

– Integrated advanced credit risk parameters, including Debt-to-Income (DTI) and Loan-to-Income (LTI) ratios, to evaluate loan eligibility and EMI affordability, presenting insights through an intuitive UI. VH24-DocMind (RAG System) - Python, Vector DB, Ollama, Streamlit, FastAPI Repository

– Architected a production-ready Retrieval-Augmented Generation (RAG) framework, integrating advanced vector databases for semantic similarity retrieval to ground LLM responses in verifiable custom data.

– Containerized application microservices utilizing Docker and deployed high-performance inference endpoints via FastAPI, optimizing system architecture for low-latency query resolution and effectively eliminating model hallucination rates.

Stock Price Trend Analysis - Python, Pandas, Matplotlib, Seaborn, Time-Series Repository

– Processed and analyzed 5+ years of historical market data ( 50,000 data points) to model price volatility and identify critical support and resistance zones.

– Developed moving-average statistical models to detect short-term momentum shifts and generate potential algorithmic trading signals.

ACHIEVEMENTS & CERTIFICATIONS

• Competitive Programming (LeetCode): Achieved a Contest Rating of 1834 (Top 7% globally), solving 200+ algorithmic problems focusing on Data Structures, Algorithms and optimization.

• Professional Certifications: Anthropic (Model Context Protocol, AI Fluency: Framework & Foundations, AI Capabilities and Limitations) Cisco Networking Academy (Python Essential 1 & 2).

• Awards & Leadership: Awarded the Best Cadet Award by the National Cadet Corps (NCC) in 2020.

Contact this candidate