mana9512 / weatherwatch

Anomaly Detection for Climate Shifts

Greetings, fellow developers!

🚀 Software Developer | Passionate Creator | Impact Seeker

🌟 With over a year of software development experience, I thrive on turning innovative ideas into impactful projects that touch people's lives.

🛠️ From crafting elegant code to building solutions that make a difference, I'm dedicated to creating technology that drives positive change and empowers communities.

💡 Let's collaborate on projects that inspire and transform. Together, we can make a difference, one line of code at a time.

📫 Reach out to me for exciting opportunities and meaningful collaborations!

✨ Let's build a better tomorrow, today! ✨

My Projects 🌟

Cloud Computing

Cloud Native User Management System on GCP

• [Tech Stack] -

• Engineered a cloud-native user management system deployed on Google Cloud Platform (GCP).
• Implemented automated deployment of Node.js APIs using GitHub Actions, streamlining the CI/CD pipeline.
• Utilized Packer and Terraform for infrastructure as code, ensuring consistent and reproducible deployments
• Implemented unit and integration tests using Jest, Mocha, and Chai.

Machine Learning and Deep Learning

Medical Image Gen

• [Tech Stack] -

• Developed a medical image generation system using advanced deep learning techniques.
• Fine-tuned the CLIP component of a diffusion model to generate images of skin diseases from textual prompts.
• Implemented a user-friendly interface using Streamlit for easy interaction with the model.

ConvoSum (Summarizing Dialogues with Ease)

• [Tech Stack] -

• Implemented a Text Summarizer project utilizing Large Language Models (LLMs) to condense conversation into precise summary.
• Leveraged pre-trained pegasus LLM from hugging face transformers to fine tune samsum dataset.

Facial Expression Recognition with Deep Learning

• [Tech Stack] -

• Developed a deep learning model to classify facial expressions into 7 emotion categories (angry, disgusted, fearful, happy, neutral, sad, surprised).
• Utilized a Convolutional Neural Network (CNN) architecture trained on the FER-2013 dataset, achieving 63% accuracy on the test set.
• Implemented real-time emotion detection using OpenCV for face detection and the trained model for classification.

Bank Term Deposit Classifier

• [Tech Stack] -

• Developed and compared three models Logistic regression, Decision Tree and Random forest respectively.
• Performed feature selection, exploratory data analysis and model evaluation.
• Addressed imbalanced classes, showcasing the unbiased nature of Tree based classifiers.

Full Stack Projects

Stresser Website

• [Tech Stack] -

• Developed RESTful APIs using the Django framework to augment API performance and implemented personality prediction utilizing K-means clustering. Achieved an accuracy rate exceeding 80% through analysis of DASS (Depression, Anxiety, and Stress Scale) test data.

Doc-Connect

• [Tech Stack] -

• A Doctor - Patient Management Portal in which patient has ability to schedule an appointment with doctor.

⚡ Technologies

Languages

Web and Databases

Frameworks

Tools

📬 Find me at