SYED MUHAMMAD ZAWWAR ASIF

As a Data Scientist, Syed Muhammad Zawwar Asif automates critical reconciliation processes, analyzes sensor data for accuracy, and provides essential technical support, significantly enhancing operational efficiency and data integrity.

• Automated manual dip-read reconciliation using Selenium, reducing processing time by 90% and eliminating data entry errors.
• Analyzed RTG sensor data to detect discrepancies in tank metrics (level, temperature, density), ensuring accuracy in reporting and preventing potential losses.
• Provided technical support for depot operations, resolving 50+ incidents monthly and streamlining multi-bay truck loading workflows.

As a Data Science Intern, Syed Muhammad Zawwar Asif contributed to a Telecommunication Churn Prediction project, exploring emerging technologies and enhancing professional communication skills.

• Contributed to a Telecommunication Churn Prediction project during a virtual internship, applying advanced data science techniques.
• Attended weekly technical and non-technical online sessions, enhancing both domain knowledge and professional communication skills.
• Explored and utilized emerging technologies, including Llama 3.2, Plotly, Seaborn, and Scikit-Learn, for data visualization and machine learning tasks.

As a Project Management Intern, Syed Muhammad Zawwar Asif managed 5 simultaneous web development projects, ensuring on-time delivery and high client satisfaction through strategic planning and cross-functional collaboration.

• Managed 5 projects simultaneously, delivering websites on time with excellent client feedback and achieving high client satisfaction.
• Applied strong communication, problem-solving, and strategic planning skills for efficient execution and successful project delivery.
• Collaborated with cross-functional teams to optimize workflows and boost productivity, enhancing project outcomes.

As a Junior AI Developer, Syed Muhammad Zawwar Asif developed and trained machine learning models, specializing in computer vision and object detection, through advanced image annotation and processing techniques.

• Performed multiple types of image annotation using RoboFlow, creating high-quality datasets for machine learning models.
• Gained proficiency in OpenCV for image processing and OCR techniques, enhancing data extraction capabilities.
• Trained models on Google Colab, specializing in YOLO for real-time object detection for various applications.

Built a no-code automation workflow in n8n to transform raw CSV files into AI-powered insights and dynamic visualizations. Integrated Groq-Kimi for trend and anomaly detection, QuickChart for automated graph generation, and email delivery for seamless reporting, enabling rapid decision-making without manual analysis.

Developed an automated system for fuel tank dip-read reconciliation, leveraging Python and Selenium to extract and compare data, significantly reducing manual effort and errors for improved data integrity.

Designed and implemented a churn-prediction pipeline using Scikit-Learn (Random Forest, XGBoost), achieving AUC 0.85, and created a Streamlit dashboard with a LangChain RAG assistant for conversational data exploration.

Developed an e-commerce product data scraping and categorization system, leveraging Selenium for data collection and a Naive Bayes classifier for product classification with 87% accuracy.

Designed and developed responsive websites for clients, ensuring modern layouts and robust functionality using HTML, CSS, JavaScript, and PHP, leveraging AI for optimization.

Trained a multi-class YOLO model for international ID and cheque detection, achieving 90% plus classification accuracy, and automated dataset augmentation to improve model robustness.

Created an OCR pipeline using OpenCV and Tesseract to extract text from printed labels, achieving 95%+ recognition accuracy, annotated over 5,000 images with Roboflow, and designed a Pandas-based ETL process to convert raw OCR outputs into clean, structured CSV files, reducing manual entry by 80%.

Developed and trained a Pyzbar-powered barcode detection model, streamlining product ID workflows, using Roboflow for dataset annotation and Google Colab for rapid model prototyping.

Research focusing on enhancing e-commerce product discovery through Selenium-based web scraping and Naive Bayes classification for improved product categorization.