MK
Data Science
Turning data into actionable insights using cutting-edge data science tools.
Technical Skills
Technologies and frameworks I work with
Python
Go
C
C++
Pandas
Numpy
Matplotlib
Seaborn
Plotly
Tensorflow
LangChain
Scikit-learn
Keras
RAG
NLTK
SpaCy
MySQL
Vector Database
PostgreSQL
MongoDB
Git
GitHub
Flask API
Docker
Heroku
Vercel
Streamlit
Netlify
Selenium
BeautifulSoup
Jupyter Notebooks
Colab
Tech & Skills Showcase
Python & Machine Learning
- Data Analysis & Visualization (pandas, matplotlib)
- Machine Learning (scikit-learn, model evaluation)
- Deep Learning (TensorFlow / Keras - image & audio models)
- Signal / Audio processing (librosa - gunshot & sound detection)
Backend, APIs & Databases
- Flask & FastAPI (REST APIs, file uploads)
- Streamlit / Gradio for demos & realtime UIs
- Database integration: PostgreSQL, MongoDB (and ORM/SQLAlchemy)
- Deployment stacks: Gunicorn, Nginx, CI/CD (GitHub Actions)
AI, LLMs & Recommenders
- LangChain & custom LLM agents
- Personalized recommendation systems (collab + content-based)
- Embeddings, retrieval-augmented generation, evaluation pipelines
Computer Vision & Detection
- YOLO (custom training & modifications) - object detection
- OpenCV, MediaPipe & DeepFace (face/emotion/gender)
- Real-time detection features license plate, helmet, speed, alerts
Cloud, Containers & DevOps
- AWS: EC2, S3, Lambda (basic infra + storage)
- Docker & Kubernetes (containerization & orchestration)
- CI/CD: GitHub Actions; deployments to Hugging Face Spaces / cloud VMs
Tools, Testing & Productivity
- VS Code & JetBrains IDEs; Git & GitHub workflow
- Developer tools & workflows - Flask, Docker, Streamlit; web scraping & testing with BeautifulSoup and Selenium
- Pytest for unit/integration testing; performance & profiling
Focused on production-ready ML systems, realtime CV apps, and backend integrations - actively building and learning new tools.
Projects
CCTV Surveillance System
An intelligent surveillance application built with Streamlit, designed for
real-time monitoring, smart detection, and automated alerts.
(IN-PROGRESS)
✨ Key Features
- ▹Real-time CCTV feed monitoring
- ▹Face, person & vehicle detection
- ▹License plate recognition
- ▹Gunshot & scream detection (audio + video)
- ▹Fire detection & early alerting
- ▹Accident detection & incident logging
- ▹Wanted car & wanted face alert system
⚙️ Technical Implementation
- ▹YOLO-based object & face detection
- ▹CNN-based fire & accident detection (vision)
- ▹Multi-modal audio analysis for scream & gunshot detection
- ▹Streamlit with WebRTC for live camera feed
- ▹MongoDB for user data, detection history & incident logs
- ▹Exploratory Data Analysis (EDA) for model diagnostics
- ▹Real-time alerting (email/SMS/push) and dashboard
Developer Dashboard
Live GitHub & coding profile highlights - contributions, stats, trophies and quick links.
