KumarAryan

Data Analyst & AI/ML Engineer

Building intelligent models, reproducible pipelines & scalable web applications

Kumar Aryan
01

About

Final Year @ VIT University, navigating the world of Data Science and ML with a keen interest in real-world applications. Specializing in Cloud Computing with hands-on exposure to Amazon AWS & Eager to grow as a full-stack developer and build impactful, scalable tech solutions.

Adept at developing GenAI-powered applications leveraging LLMs, RAG, and Agentic AI workflows. Comfortable working across the full AI development lifecycle – from model building to deployment using Flask, Docker, and containerized environments.

Interests

  • Machine Learning & AI
  • Genomic Data Modeling
  • Data Analytics
  • Full-Stack Development
  • Competitive Programming
02

Education

Vellore Institute of Technology (VIT)

2023 – 2027
B.Tech in Computer Science & EngineeringCGPA : 8.2 / 10

DAV Centenary Public School

Class XIIPercentage % : 84
2022
Class XPercentage % : 97
2020

03

Experience

01

Machine Learning Summer School

AmazonApprenticeship
August 2025Remote
  • Implemented supervised, unsupervised, and complex machine learning algorithms using Python libraries such as scikit-learn, NumPy, and pandas.
  • Developed machine learning models for diverse projects using advanced algorithms and tools, achieving accuracy rates exceeding 95%.
Machine LearningPythonscikit-learnNumPypandasAlgorithms
04

Projects

Indian Genome Atlas

124K Markers
Indian Genome Atlas screenshot

First-of-its-kind scalable qpAdm wrapper enabling a fully reproducible admixture pipeline breaking the ancestral population, analyzing 124K SNP markers across 100+ individuals using AADR dataset

  • Python
  • admix tools
  • AADR
  • qpAdm
  • Seaborn
  • PLINK

PhishGuard

98.5% Acc
PhishGuard screenshot

Machine learning–based phishing detection system with a Chrome extension, leveraging Random Forest, XGBoost, and SVM to classify and detect malicious URLs, achieving 98.5% accuracy and 97.8% precision

  • Flask
  • XGBoost
  • SVM
  • scikit-learn
  • Chrome Extension

SkillMapper

AI Powered
SkillMapper screenshot

A full-stack MERN-based learning platform leveraging LangChain and Gemini API to generate personalized, role-specific learning roadmaps based on users’ skills, career goals, and available learning time

  • React.js
  • Node.js
  • LangChain
  • Gemini API
  • MongoDB
  • JWT

Project Synapse

Prompt Engg.
Project Synapse screenshot

An autonomous agentic system for last-mile delivery coordination, orchestrating a ReAct-style reasoning loop via LangGraph to resolve real-time disruptions beyond static rule-based logic. Used a Hugging Face-hosted LLM for dynamic tool invocation

  • Flask
  • Agentic Framework
  • Hugging Face API
  • LangChain

FakeReview Sentry

ML Classifier
FakeReview Sentry screenshot

Multi-signal ensemble classifier fusing NLP sentiment with reviewer behavior, engineering sentiment polarity, posting frequency, and verified-purchase ratio to flag suspicious reviewer patterns

  • Python
  • Scikit-learn
  • XGBoost
  • NLP
  • SHAP
  • Flask
  • Docker
05

Tech Stack

PythonPython
SQLSQL
JavaJava
TensorFlowTensorFlow
PyTorchPyTorch
KerasKeras
Scikit-learnScikit-learn
MatplotlibMatplotlib
SeabornSeaborn
qpAdmqpAdm
LLMsLLMs
RAGRAG
Agentic AIAgentic AI
NLPNLP
GANsGANs
OpenAI APIsOpenAI APIs
GitGit
GitHubGitHub
DockerDocker
KubernetesKubernetes
Jupyter NotebookJupyter Notebook
KaggleKaggle
admix toolsadmix tools
PythonPython
SQLSQL
JavaJava
TensorFlowTensorFlow
PyTorchPyTorch
KerasKeras
Scikit-learnScikit-learn
MatplotlibMatplotlib
SeabornSeaborn
qpAdmqpAdm
LLMsLLMs
RAGRAG
Agentic AIAgentic AI
NLPNLP
GANsGANs
OpenAI APIsOpenAI APIs
GitGit
GitHubGitHub
DockerDocker
KubernetesKubernetes
Jupyter NotebookJupyter Notebook
KaggleKaggle
admix toolsadmix tools

Programming Languages

  • Java
  • Python
  • SQL

Frameworks & Libraries

  • TensorFlow
  • PyTorch
  • Keras
  • Scikit-learn
  • Matplotlib
  • Seaborn
  • qpAdm

AI/ML & GenAI

  • LLMs
  • RAG
  • Agentic AI
  • NLP
  • GANs
  • OpenAI APIs

Developer Tools

  • Git
  • GitHub
  • Docker
  • Kubernetes
  • Jupyter Notebook
  • Kaggle
  • admix tools
06

Achievements

MP Police CyberSafety Hackathon

Finalist (Top 5 Teams)

Top 5 Finalist

Recognized as a Finalist in the CyberSafety Hackathon organized by MP Police, securing a spot among the top 5 teams for building innovative security solutions.

Rating: 1300
CODEFORCES

Top 30% of active global participants in competitive programming on Codeforces.

University of Michigan

Applied ML in Python

Applied ML in Python
View Certificate
Amazon AWS

AWS Certified Cloud Practitioner

AWS Certified Cloud Practitioner
View Certificate
Oracle

Certified Data Science Professional

Certified Data Science Professional
View Certificate
Data Structures

300+ DSA Problems Solved

Solved 300+ algorithms and data structures problems with focus on optimization and problem solving.

GitHub

Active Contributions

in the last year

GitHub Contributions
07

Socials

08

A Personal Note

Thanks for making it all the way to the end! I have a little something for you. I hope it resonates with you in some way, and maybe even gives you a little boost of motivation or perspective.

>...
0:00
0:00