Hello, I'm

Vishnu Kumar Sharma

Aspiring Data Scientist | Machine Learning | GenAI | MLOps

Turning messy data into working models — and working models into usable products

View Projects Contact Me

About Me

I'm an aspiring Data Scientist who enjoys turning messy data into working models — and working models into usable products.

Comfortable across the ML lifecycle — EDA, feature engineering, model training and evaluation. Learning to build production-style ML — modular pipelines, experiment tracking, containerization and APIs.

Increasingly exploring GenAI & LLM-powered applications — agents, tool-calling, natural-language interfaces. Learning by building — internships, guided specializations, and self-directed projects.

Always happy to talk data science, MLOps, or GenAI — open to collaborating.

ML lifecycle: EDA → feature engineering → training → evaluation
Production ML: pipelines, experiment tracking, containerization, APIs
GenAI & LLMs: agents, tool-calling, RAG, fine-tuning (LoRA/QLoRA)
Cloud & MLOps: AWS, Azure, Docker, Kubernetes, MLflow, Databricks
Learning by building: internships, specializations, self-directed projects

Tech Stack

Languages & Notebooks

Python
SQL
Jupyter

Machine Learning & Deep Learning

scikit-learn
Pandas
NumPy
TensorFlow
PyTorch

GenAI, LLMs & NLP

LangChain
LangGraph
Hugging Face
OpenAI
NLTK
Ollama
llama.cpp

Computer Vision

OpenCV
MediaPipe

Time Series

Statsmodels
ARIMA
Prophet

Data Visualization & BI

Matplotlib
Seaborn
Power BI

MLOps & Deployment

MLflow
Docker
Kubernetes
Flask
FastAPI
Streamlit
SQLite

Cloud & Data Platforms

AWS
Azure
Databricks

Tools

Git
GitHub
VS Code

Featured Projects

🍷 End-to-End Wine Quality Predictor

Full ML pipeline (ingestion → validation → training → deployment) predicting wine quality, with MLflow + DAGsHub experiment tracking and a Dockerized Flask API

Pythonscikit-learnMLflowDockerFlask

Code

📈 Marketplace Analysis Agent

GenAI market-research copilot that answers natural-language stock questions using live price data, fundamentals, risk metrics and news

LangChainOpenAI GPTStreamlitEODHD API

Code

🌍 Carbon Emission Forecasting

Predicts country-wise CO₂ emissions per capita with a Random Forest model (RFECV feature selection) and an interactive Streamlit demo

scikit-learnStreamlitRandom ForestRFECV

Code

❓ Quora Duplicate Question Detection

NLP pipeline that flags duplicate Quora question pairs via text preprocessing, feature engineering and classical ML

NLTKscikit-learnNLP

Code

🌳 Tree Species Classification

CNN-based image classifier that identifies tree species to support ecological and biodiversity research

Deep LearningCNNComputer Vision

Code

💰 Spendly — Expense Tracker

Full-stack personal finance tracker with authentication, CSRF/XSS hardening, CSV export and an automated pytest suite

FlaskSQLitepytest

Code

📌 More experiments — classical algorithms, NLP summarization, GFG 21-Day GenAI & Deep Learning challenges — on my repositories page.

Currently Leveling Up Through

vLLM — fast, production-grade LLM inference & deployment
Agentic AI, AI Agents & Multi-Agent Systems
LLMs & Transformer Architecture
Production-Grade AI Engineering & System Design
LLM Deployment & Model Serving (vLLM, Ollama, TGI, llama.cpp)
LLM Fine-Tuning (LoRA, QLoRA, PEFT, SFT)
Production Training Pipelines & MLOps for LLMs
Building LLMs from Scratch
Production-Level AI & GenAI Projects

Contact

Got an opportunity or a problem to solve? I'd love to hear from you — always open to feedback, collaboration, or a good data science conversation.

LinkedIn Email