Technical Skills & Toolkit

An overview of technical competencies, frameworks, infrastructure tooling, and algorithms developed across academic research and applied machine learning engineering.

Skills Matrix

• Deep Learning & AI: PyTorch, TensorFlow / Keras, Hugging Face Transformers, Vision Transformers (ViT, Swin), BERT / DeBERTa / RoBERTa, CLIP, Cross-Modal Attention, CNNs (ResNet, VGG), OpenCV, Scikit-learn, XGBoost.
• MLOps & Cloud: DVC (Data Version Control), MLflow (Experiment Tracking & Model Registry), Docker (Containerization), FastAPI, REST APIs, Streamlit, Git, GitHub Actions CI/CD, AWS (EC2, S3), DagsHub.
• Languages & Core: Python, C++, C, SQL (PostgreSQL, MySQL), JavaScript (ES6+), HTML5/CSS3.
• Core Competencies: Computer Vision, Natural Language Processing, Multimodal Gated Fusion, Data Structures & Algorithms (250+ solved challenges), GATE Qualified (CS 514), Explainable AI (SHAP), Statistical Hypothesis Testing.

Development Principles

• Reproducibility First: Strict seed management, deterministic data splits, and DVC pipeline definitions.
• Low-Latency Deployment: Asynchronous FastAPI microservices with optimized torchscript / onnx inference models.
• Continuous Experiment Tracking: Comprehensive metric and artifact logging with MLflow and automated regression checks.