Skip to content
View harmandeep2993's full-sized avatar
📚
Learning ML & Python | Open to opportunities
📚
Learning ML & Python | Open to opportunities

Block or report harmandeep2993

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
harmandeep2993/README.md

Hi there 👋, I'm Harman

Machine Learning Practictioner · Data Scientist · MLOps · Generative AI Systems

LinkedIn Portfolio Tableau

👋 About Me

Machine Learning Engineer focused on building end-to-end AI systems from model development to production deployment. Experienced in Computer Vision, NLP/LLMs, and MLOps, with strong emphasis on scalable and reproducible machine learning pipelines.

Based in Germany and actively seeking opportunities in AI Engineering, Machine Learning Engineering, and Computer Vision roles.

🧠 What I Build

  • End-to-end Machine Learning systems (training → API → deployment)
  • Real-time Computer Vision applications (edge inference & APIs)
  • LLM-powered applications (RAG, embeddings, multi-provider LLM systems)
  • Production ML pipelines with Docker, FastAPI, AWS, and CI/CD

🚀 Featured Projects

  • Built a real-time emotion detection system using EfficientNet-B0 for classification and MediaPipe for face detection, exposed through a FastAPI REST API for inference, and implemented CLAHE-based image preprocessing to improve robustness in low-light conditions.
  • Impact: Real-time inference pipeline optimized for CPU deployment
  • Built a collaborative filtering recommendation system using SVD on the MovieLens-1M dataset, improved performance by reducing RMSE from 2.52 to 0.965 through iterative MLflow experiments, and deployed the model as a scalable REST API using FastAPI on AWS ECS (Fargate) with a complete MLOps pipeline including CI/CD, data versioning, and automated testing.
  • Impact: End-to-end production ML system with cloud deployment
  • Built an AI-powered resume–job matching system using Sentence Transformers and LLMs for semantic similarity scoring, combined embeddings with keyword-based ranking for hybrid relevance, implemented multi-LLM routing (OpenAI, Groq, Ollama) for flexible inference, and ensured structured outputs using Pydantic schema validation via a FastAPI service deployed with Docker.
  • Impact: Intelligent skill-gap analysis and job matching engine

🧪 Other Projects

🛠️ Technical Skills

Programming Languages & Tools

  • Python, SQL

Machine Learning & AI

  • Pandas, Numpy, Scikit-learn, XGBoost, TensorFlow, PyTorch
  • NLP, LLMs, RAG, Sentence Transformers, Hugging Face
  • Computer Vision, OpenCV, MediaPipe

MLOps & Cloud Deployment

  • FastAPI, Docker, MLflow, DVC, GitHub Actions, REST APIs
  • AWS (S3, Lambda, API Gateway, ECS, ECR)

📌 Focus Areas

  • Production Machine Learning (MLOps)
  • Computer Vision systems
  • Generative AI applications (LLMs, RAG)

📫 Open To Opportunities

Actively seeking roles in Machine Learning Engineering, AI Engineering, and Computer Vision across Germany and Europe.

Pinned Loading

  1. JobFitAI-public JobFitAI-public Public

    This is the public showcase for JobFitAI. The full source code lives in a private repository while the project is under active development.

    HTML

  2. DrowsyGuard DrowsyGuard Public

    Real-time driver drowsiness detection using MediaPipe and OpenCV. Detects eye closure and yawning from webcam feed with audio alerts and session logging.

    Python

  3. CineMatch CineMatch Public

    End-to-end movie recommendation system with collaborative filtering, experiment tracking with MLflow, REST API with FastAPI, containerization with Docker, and CI/CD deployment to AWS.

    Python

  4. EmoDetector EmoDetector Public

    Real-time facial emotion detection pipeline using MediaPipe BlazeFace and EfficientNet-B0. REST API with FastAPI. Runs on CPU. Docker ready.

    Python

  5. customer-churn-predictor customer-churn-predictor Public

    Python

  6. image-classification-cnn-transfer-learning image-classification-cnn-transfer-learning Public

    Jupyter Notebook