Innovative Machine Learning Solutions for Real-World Challenges

Empowering Technology Through Advanced AI and Engineering Skills

Explore

AI Solutions

I develop intelligent systems that solve real-world problems by combining deep learning, data understanding, and software engineering

Burn Degree Classification

  • Objective: Detect the severity of burn injuries (1st, 2nd, 3rd degree) from medical images using deep learning.

  • Tech Stack: TensorFlow, EfficientNetB3, Google Cloud Run

  • Results: Achieved 92% accuracy on validation data.

  • My Role:

    • Fine-tuned EfficientNetB3 model

    • Performed evaluation using Precision, Recall, and F1-Score

    • Deployed model via Flask API on Google Cloud

Smart Recipe Generator

  • Objective: Recommend recipes based on the ingredients currently available in the user’s fridge.

  • Tech Stack: TensorFlow, NLP, Firebase, Streamlit

  • Features:

    • Ingredient detection from user input

    • Recipe matching using similarity scoring

    • Real-time recipe suggestions with simple UI

Customer Churn Prediction Dashboard

Objective:
Analyze customer behavior and predict churn (customer dropout) based on historical transactional and demographic data to support business decision-making

Tech Stack:
Pandas, Scikit-learn, XGBoost, Seaborn, Plotly, Streamlit

Results:
Achieved 84% accuracy and AUC-ROC score of 0.91 in predicting customer churn.Performed data cleaning, feature engineering, and missing value imputation on >10,000 customer records.

  • Conducted Exploratory Data Analysis (EDA) using Seaborn & Plotly to identify trends and high-risk churn segments.

  • Built classification models (Logistic Regression, Random Forest, XGBoost) and compared performance using cross-validation.

  • Created an interactive dashboard using Streamlit that visualizes:

    • Churn prediction results

    • Customer segmentation by churn risk

    • Key features affecting customer retention

Github
Github

Work Experience

AI Developer & Full Stack Developer

I held a full-time role as an AI Developer and Full-Stack Engineer, responsible for the end-to-end development, deployment, and optimization of sophisticated AI applications and their supporting web interfaces. My focus was delivering high-performance, robust, and scalable solutions from model training to production.

  1. Led the development of advanced AI models, including Object Detection and Fraud Detection engines, integrated directly into business workflows.

  2. Designed and built full-stack applications using Vue.js for the frontend, Laravel for scalable backend APIs, and SQL for database management.

  3. Developed and deployed custom Large Language Model (LLM) solutions, leveraging platforms such as OpenAI, Gemini, and Ollama, managing various cloud and local environments.

  4. Ensured seamless integration between complex AI models and the production environment, focusing on deployment efficiency and real-time performance optimization.

  5. Managed the entire development lifecycle, utilizing Git for version control and collaborating closely with cross-functional teams in an agile environment.

  6. Advised on system architecture and code quality through active participation in design sessions and stringent code reviews.

Github

Work Experience

Website Developer

As a Frontend Developer Intern, I was responsible for assisting in the development and enhancement of user-facing features for web applications. My role focused on creating responsive, interactive, and visually appealing interfaces that provide seamless user experiences.

1. Collaborated with UI/UX designers to translate wireframes and design mockups into high-quality code.
2. Developed reusable components using HTML, CSS (Tailwind/Bootstrap), and JavaScript frameworks (e.g., React.js or Vue.js).
3. Assisted in debugging and optimizing front-end performance for better speed and scalability.
4. Participated in code reviews and team meetings using version control (Git) and agile methodologies.
5. Integrated APIs and backend services into the frontend architecture.
6. Ensured cross-browser compatibility and responsive design across devices.

Github

Get in Touch

I'm eager to connect and collaborate on innovative AI solutions. Let's discuss how technology can solve real-world challenges together.