❯ whoami
AI Engineer | Data Scientist (in training)

❯ neofetch
-------------------------
OS: macOS
Host: MacBook
Kernel: Darwin
Shell: zsh
CPU: Apple Silicon
Editor: VS Code, Jupyter, IntelliJ IDEA
Uptime: Shipping ML pipelines daily

❯ pwd
/Users/dhruv/github

❯ ls
machine-learning  computer-vision  llm-agents  mlops  data-eng  cloud

❯ cat about.txt
3rd-year B.Tech CSE (AI & DS), Graphic Era Deemed University
Building production-grade ML systems, not just notebooks
Currently deep in MLOps: training -> tracking -> serving -> monitoring

❯ cat skills.txt
Languages     Python | SQL | Java | C/C++
ML/DL         XGBoost | PyTorch | TensorFlow/Keras | YOLOv8 | Optuna | SHAP
LLM/Agents    RAG pipelines | Multi-agent systems | LLM tooling
MLOps         MLflow | FastAPI | Docker
Data Eng      Pandas | ETL pipelines | SQL
Cloud         AWS (EC2, S3, IAM, Lambda, API Gateway)
Web/Deploy    Streamlit | Vercel | REST APIs
Tools         Git | GitHub | Linux

❯ ls projects/
AlertVision-Violence-Detection-System  FocusGuard  churn-prediction-mlops

❯ cat projects/AlertVision-Violence-Detection-System/README.md
Real-time violence detection from video (ResNet50 + LSTM)
Frame sampling -> CNN feature extraction -> temporal classification
Streamlit UI, configurable sensitivity, automated email alerts
Trained on RWF-2000 CCTV dataset | Live on Hugging Face Spaces
stack: Python, TensorFlow/Keras, OpenCV, Streamlit
repo:  github.com/Dhruv-Sharma29/AlertVision-Violence-Detection-System

❯ cat projects/FocusGuard/README.md
Privacy-first developer productivity tracker (terminal-based)
Time-decayed focus score from local activity
Flags procrastination via weighted compound signals
Zero telemetry | AES-256 encrypted local storage
Opt-in AI coaching via local or hosted LLMs
stack: Python, CLI, SQLCipher
repo:  github.com/Dhruv-Sharma29/FocusGuard

❯ docker ps
CONTAINER ID   IMAGE                STATUS        PORTS
b3f9a2c1e0     churn-api:fastapi    Up 3 hours    8000->8000
e7d4c8a2f1     mlflow-tracking      Up 3 hours    5000->5000
c1a9d3e7b5     evidently-monitor    Up 3 hours    8080->8080

❯ git log --oneline
9f2a7c1  Add SHAP explainability to churn model
e4b8d3a  Deploy FastAPI backend to AWS Lambda
b1c6f92  Integrate Evidently AI for drift monitoring
a7d3e8c  Tune XGBoost with Optuna hyperparameter search
f2c9b4e  Set up MLflow experiment tracking
d1e5a3c  Ship AlertVision violence detection system
c8b2f91  Build FocusGuard CLI + focus-score engine

❯ cat contact.txt
linkedin  : linkedin.com/in/dhruv-sharma-ds28
github    : github.com/Dhruv-Sharma29