Chandrasekaran Subramanian Ravichandran

A Data Analyst with an experience in compiling, processing, and analyzing complicated Data. I am trained in large-scale Data Analytics with a focus on Big Data, Cloud Computing. I am highly proficient in Cloud solutions such as AWS (Sagemaker, S3, EC2) and IBM Watson Studio (AutoAI, SPSS), and have practical expertise with TensorFlow, PyTorch, Keras, Scikit-learn and Database Management. I have a solid foundation in Data governance and Data integrity. Successfully built relationships and developed solutions to corporate issues. I have a strong passion for Data-driven solutions.

ETL, Data (Mining, Wrangling, Warehousing), AWS (S3, EC2, Sagemaker), Python, MS Excel,  JIRA, Git Version Control, NLP, Deep Learning, MLOps.

 1. Time-Series Based Telecom Call Drop Prediction December 2021 - May 2022

• Developed a machine learning model to predict call drops using time-series data and external factors related to Base Transceiver Station (BTS) in a telecommunication network.
• Preprocessed and cleaned a large dataset comprising signal strength, network load, user mobility, and various environmental factors (e.g., weather conditions, infrastructure issues).
• Engineered features from raw BTS data to create effective predictors for the model, such as signal-to-noise ratio (SNR), Received Signal Strength Indicator (RSSI), and network congestion metrics.
• Applied advanced time-series analysis techniques, including ARIMA, SARIMA, and LSTM (Long Short-Term Memory) neural networks, to forecast call drop rates.
• Tuned model parameters using techniques such as Grid Search and Cross-Validation to optimize performance metrics, including RMSE (Root Mean Square Error), MAPE (Mean Absolute Percentage Error), and AUC-ROC (Area Under the Receiver Operating Characteristic curve).
• Visualized results using tools like Matplotlib and Seaborn to provide clear and interpretable insights into call drop trends and model performance.
• Implemented a user-friendly dashboard using Streamlit to enable real-time monitoring and alerting based on the model's predictions.
• Managed code versioning and project documentation through Git, Jira and made the entire project publicly available for community contributions and reviews.

GitHub Repository: https://github.com/vimaleshraja/calldrop

2.  Architectural Chatbot with IBM Watson Studio January 2021 - March 2021

• Platform: Watson Studio
• Deployment Platform: Facebook Messenger
• Conceptualized and developed an interactive architectural chatbot using IBM Watson Studio, tailored to provide insights and information related to architectural concepts, designs, and queries.
• Seamlessly integrated the chatbot into Facebook Messenger, enabling wide accessibility and user engagement on a popular platform.
• Implemented natural language processing (NLP) capabilities for accurate query understanding and response generation, offering users a conversational and informative experience.
• Regularly monitored user interactions and collected feedback to iterate and refine the bot for better user experience and accuracy.

Link: https://www.facebook.com/profile.php?id=100069780467739