Vedant Marathe

About Vedant Marathe

I’m an AI & Data Science enthusiast with a strong Computer Engineering foundation, focused on turning raw data and complex systems into clear insights, working models, and practical AI solutions. My learning has always been hands-on — every concept I studied was immediately converted into an experiment, a working pipeline, or a real project.During my AI & Data Science specialization, I worked directly with real-world datasets, built end-to-end ML systems, and developed an engineering mindset for how data moves, how models behave, and how scalable solutions are designed. My Computer Engineering background in algorithms, data structures, operating systems, and networks strengthened my ability to reason about performance, logic, and system reliability behind every model.Across the program, I consistently transformed theory into measurable outcomes:• Practiced EDA, NumPy, and Pandas on real datasets — cleaning data, resolving anomalies, exploring feature relationships, and forming hypotheses that guided decisions.• Built regression, classification, and clustering models after each ML module, focusing on feature design, evaluation, and translating predictions into business-ready insights.• Delivered practical projects such as sentiment analysis, customer churn prediction, GPT fine-tuning for Indian recipe generation, and a fraud detection capstone centered on risk-pattern discovery and explainability.• Advanced into Deep Learning by training ANN, CNN, and RNN architectures for image, sequence, and structured data, evaluating them under real-world constraints such as performance, overfitting, and data imbalance.• Applied Generative AI by fine-tuning GPT-2 and FLAN-T5 using PEFT/LoRA and deploying interactive AI applications through Gradio, combining both engineering depth and a focus on usability.Through these experiences, I learned not only how to build models but how to interpret them, identify data issues, design stronger features, and communicate insights in a way that enables clear decision-making for technical and non-technical teams.What excites me most about this field is the blend of analytical discipline, creativity, and impact. Whether detecting fraud faster, simplifying monitoring systems, segmenting customers intelligently, or automating sentiment understanding — the goal is always to deliver clarity, improve decisions, and reduce effort.I’m actively seeking opportunities as where I can apply my hands-on experience, engineering mindset, and commitment to practical outcomes to solve meaningful problems

Experience

1. Ai & Data Science Trainee Applied Learning Program

   Boston Institute of Analytics

   Jul 2025 — Present · Mumbai, IN

   During this hands-on program, I worked on real datasets, built end-to-end ML systems, and applied every concept through structured practicals and real projects.• Built strong analytical foundations by working with Excel and SQL datasets—cleaning data, resolving inconsistencies, writing optimized queries, and identifying patterns that supported business decision-making.• Applied Python concepts through continuous implementation—using NumPy, Pandas, and visualization libraries to explore data, engineer features, and transform raw information into clear insights.• Strengthened statistical thinking through hypothesis testing, regression diagnostics, and distribution analysis to validate assumptions and support data-driven conclusions.• Developed regression, classification, and clustering models after each ML module—focusing on feature design, evaluation, and translating model outputs into actionable interpretations.• Delivered structured projects such as sentiment classification, customer churn prediction, and a fraud detection capstone—emphasizing high-quality preprocessing, risk-pattern discovery, and explainability.• Advanced into deep learning by building ANN, CNN, and RNN architectures for image, sequence, and structured data, evaluating performance under practical constraints.• Applied Generative AI by fine-tuning GPT-2 and FLAN-T5 using PEFT/LoRA, and deployed interactive AI applications via Gradio.• Improved communication skills by translating complex model behavior into clear insights that help non-technical users understand risk drivers, trends, and model outputs.

Education

• Savitribai Phule Pune University

  Bachelor of Engineering - BE

• Global Public school & Jr. College

  HSC