Professional Experience

My Career Journey:

• Junior Data Engineer (I.T Services | Digital & Product Solutions)
  • April 2023 - Present
  • Agile Datapro [Microsoft ISV Partner], Campbell, CA
• IP Product Project (Azure, Python, BlobStorage, ChatGPT 4 Turbo Key API, AI/ML, Jupyter Notebooks, Outlook, Pandas, REST, Excel):
  • Web scrapped several sites using a Python script with Selenium API, efficiently extracting 250k resumes from Outlook, achieving a consistent average of 100 job listings per minute, and collecting an extensive dataset of over 20k job listings weekly from various websites.
  • Automated data extraction processes, resulting in saving of over 80 hours each month, significantly improving productivity and data collection rates.
  • Organized and structured large amounts of data in XLSX format; encompassing attributes such as job titles, company names, locations, detailed job descriptions, and salary information.
  • Ensured quality data extraction and consistency, reducing data preparation time by 70%, facilitating efficient data analysis, and reporting.
  • Implemented data storage procedures, efficiently saving collected data to XLSX files, including a system that dynamically checks and adds missing headers to enhance data management efficiency and eliminate potential errors.
  • Created data transfer pipelines to Blob Storage, delivering a 85% reduction in manual data entry errors.
  • Conducted data preprocessing, systematically removing duplicates, and efficiently categorizing data by industry buckets, enhancing data quality by 80%.
  • Parsed and extracted critical information, including names, job titles, emails, resume contents, and top 5 skills from each resume to train the ML model and in order to generate score cards.
  • Automated scorecard generation, which resulted in 50% reduction in manual effort and saving approximately 15 hours per project cycle.
  • Tailored each score card with personalized information, including names, skills, and skill percentages, maintaining an accuracy rate of 95% in reflecting candidate profiles and skills.
  • Assisted in training the ML model to provide job recommendations and candidates, streamlining the hiring process and improved candidate matches.
    
    
• Telecom Project (GCP, Python, SQL, PostgreSQL, PowerBI, Docker, ChatGPT 3.5 Turbo Key API, AI/ML, Jupyter Notebooks, Flask):
  • Developed predictive models for a major telecom client to predict cost, utilization or social health, and deployed them on Google Cloud Platform.
  • Utilized client staging environment to extract consumer data in JSON format for analysis and processing.
  • Integrated ChatGPT API to the GCP on a PowerBI dashboard. Allows users to ask the chatbot about their multi-tenant dashboard content, issues, etc.
  • Built the chatbot using Microsoft Virtual Agent (interface for chatbot), and that Agent had an automated flow using Power Automate.
  • Conducted and drove the adoption of enterprise-wide analytics to support strategic execution using prescriptive and predictive analytics with a specialty in the Finance, Quality and Utilization domains.
  • Created visually appealing dashboards for data analysis and visualization using BI and PostgreSQL.
  • Tailored the dashboards to meet client requirements and presented complex data sets in a clear and concise manner.
  • Developed interactive, meaningful and user-friendly dashboards to enable quick identification of key insights and trends.
  • Ensured scalability of dashboards to accommodate additional data.
  • Compiled hyperparameter tuning pipelines for AI/machine learning and deep learning models. The framework enabled the company to train and deploy machine learning models 2 times faster, resulting in a 25% improvement in prediction accuracy.
    
    
    
• AIoT Project (Python, Raspberry PI, OpenCV, Tensorflow, Keras):
  • Designed an AIOT solution to an client that created an online platform to teach programming in the classroom for over 50 colleges in India.
  • Developed object detection and recognition modules for a manufacturing facility, resulting in a 40% reduction in manual labor and a 20% increase in operational efficiency.
  • Obtained data sets off Kaggle, and GitHub repositories to train the model.
  • Created an ETL pipeline using Python, and preprocessed the data to find null values.
  • Collaborated on the integration of IR sensors for real-time object recognition and tracking, streamlining manufacturing processes and improving productivity.
  • Assisted in the design and implementation of an ultrasonic sensor module for precise object placement, achieving a 15% reduction in errors.
  • Utilized advanced data science techniques to analyze sensor data, leading to a 25% reduction in equipment downtime.
  • Trained object detection models using Keras and TensorFlow, enabling the system to detect objects and display their coordinates, colors, and shapes accurately.
  • Utilized data science techniques to analyze sensor data, identifying patterns and anomalies for predictive maintenance, resulting in a 15% decrease in equipment downtime.
  • Performed thorough testing and validation of the sensor modules, ensuring accuracy, reliability, and seamless integration into the clients platform.
    
    
• Data Science and Machine Learning Intern (Ascend Technology Inc.)
  • Jan 2022 - May 2022
  • 1879 Lundy Ave STE 289, San Jose, CA 95131
  • Research Mentors: Dr. Mokhtar Sadok, Dr. Indranil Mukhopadhyay, Dr. Mohammad Akram Hossain
  • GitHub Repository: NeuroScan
• Project Title: “NeuroScan” – Machine Learning Application for Brain Tumor Detection
• Technologies: Python, TensorFlow, Keras, OpenCV, Jupyter Notebooks
  • Collaborated with Dr. Mokhtar Sadok, Dr. Indranil Mukhopadhyay, and Dr. Mohammad Akram Hossain to develop “NeuroScan,” a machine learning application designed for brain tumor detection.
  • Conducted extensive research on medical imaging datasets, including MRI scans, to identify features and patterns indicative of brain tumors.
  • Utilized machine learning algorithms such as Convolutional Neural Networks (CNNs) and Support Vector Machines (SVMs) with Python and TensorFlow to analyze and classify brain images, achieving high accuracy rates.
  • Led efforts in preprocessing and cleaning datasets to ensure robust and accurate models for tumor detection.
  • Worked closely with medical professionals to validate and refine the model’s performance, incorporating feedback to enhance precision and sensitivity.
  • Presented the project’s progress and findings in departmental seminars, showcasing technical skills and the ability to communicate complex concepts effectively.
• Citations:
  • Sadok, M. (2021, August 8). Artificial Intelligence: A paradigm shift in the pharmaceutical industry - use case of cancer detection. Digitale Transformation - jetzt die Chancen aktiv nutzen! https://www.strategy-transformation.com/artificial-intelligence-a-paradigm-shift-in-the-pharmaceutical-industry-use-case-of-cancer-detection/