Dux is a ML Engineer with 4+ years of experience in delivering innovative solutions across machine learning, natural language processing, and computer vision. Vetted by Witarist and ready to join your team within 48 hours.
Developed an intelligent, multilingual help bot, enhancing user experience and accessibility across multiple languages.
Successfully created and deployed a GANs model for facial identity transformation on a web platform.
Delivered predictive models for stock price forecasting with reasonable accuracy, providing actionable insights for investors.
Forecasted house rental prices with accuracy, leveraging Facebook Prophet to capture seasonal trends.
Utilized cross-validation techniques to ensure robustness and reliability of predictive models in financial and real estate domains.
Built Multilingual Help Bot — intelligent real-time support answering queries in English + Arabic with context-aware responses using RASA + Python + PyTorch + NLP + LLM + Hugging Face Arabic transformers
Built Facial Identity Transformation Using GANs system generating new facial images while maintaining original identity for realistic transformations on web platform using Tensorflow + GANs + Python + Flask + MongoDB
Built Stock Price Prediction predictive model forecasting future stock prices for stock-market trends + seasonal patterns + investor decision-making using Python + Facebook Prophet + Pandas + Matplotlib + NumPy
Built House Rent Forecasting model predicting future house-rental prices for real-estate investors + stakeholders using Facebook Prophet seasonal patterns + trends with Python + Pandas + Matplotlib + NumPy
Overview: Developed an intelligent, multilingual help bot providing real-time support and answering user queries in multiple languages. Responsibilities: Created a multilingual help bot for chat and voice with English and Arabic support. Used RASA for chatbot development and Hugging Face Arabic transformers for language processing. Incorporated LLM capabilities and is currently integrating ChatGPT for advanced functionality.
Key outcomes:
Enhanced user experience and accessibility by providing support across multiple languages.
Improved user satisfaction through accurate, context-aware responses.
Overview: Developed a system for facial identity transformation using Generative Adversarial Networks. Responsibilities: Specialized in generating new facial images using GANs for realistic transformations. Designed network models for attribute manipulation in image processing. Deployed the trained GAN model for user-specified transformations on a web platform using Flask.
Key outcomes:
Successfully generated new facial images while preserving original identity.
Deployed a functional GAN model for user-specified transformations.
Overview: Developed a predictive model for forecasting future stock prices using Facebook's Prophet model. Responsibilities: Acquired and cleaned historical stock price data from reliable financial sources using Pandas and NumPy. Utilized Facebook Prophet to train on historical data, capturing trends and seasonality. Generated forecasts and evaluated the model's performance, including creating visualizations with Matplotlib and Power BI.
Key outcomes:
Successfully predicted future stock prices with a reasonable degree of accuracy.
Provided insights into market trends and seasonal patterns, facilitating informed decision-making.
Overview: Forecasted future house rental prices using the Facebook Prophet model. Responsibilities: Gathered and processed historical house rent data from real estate websites and databases using Pandas. Applied the Prophet model to train on rental price data, capturing underlying trends and seasonal effects. Produced forecasts and assessed model performance, developing visualizations with Matplotlib and Power BI to compare actual and predicted values.
Key outcomes:
Achieved accurate predictions for future house rental prices.
Effectively captured seasonal patterns and trends in the rental market.
DUX
AWS Sagemaker