Vivek Parihar Email & Phone Number

Bengaluru, Karnataka, India

Noida, Uttar Pradesh, India

• Chat Log Mining Tool Development: Designed and implemented a Chat Log Mining tool that leverages LLM and BERT to analyze group chat data, providing predictive insights and solutions for previously discussed issues.• Customer Cross-Sell Prediction: Conducted comprehensive exploratory data analysis, feature engineering, and model development for predicting cross-sell opportunities, enhancing customer insights and retention strategies.• Model Deployment and Scaling: Streamlined model deployment using Flask, Docker, and Kubernetes, ensuring efficient, scalable solutions for production environments.• Production Services on Linux: Successfully deployed data science services on Linux production servers with a fully integrated UI, optimizing performance and usability for end-users.Technical Skills:• Programming: Python, SQL• Machine Learning: Scikit-Learn, TensorFlow, BERT, LLMs• Data Analysis & Visualization: Pandas, NumPy, Matplotlib, Seaborn• Natural Language Processing (NLP): BERT, Text Mining, Chat Log Analysis• Feature Engineering & Selection: Feature extraction, data preprocessing techniques• Model Deployment: Flask, Docker, Kubernetes• Version Control: Git• Cloud & Server Management: Linux (for production deployment), container orchestration• API Development: REST APIs, Flask

•Leadership in ML Pipeline Development: Led a team to build a comprehensive pipeline, covering data extraction, machine learning model development, and seamless deployment into production, enhancing pipeline efficiency and team productivity.•Advanced Text Analysis for Client Insights: Spearheaded an advanced NLP-based tool to analyze and compare complex text data, providing deeper insights and value to clients’ data.•Predictive Model Optimization: Enhanced model performance with data preprocessing and feature engineering, consistently improving model accuracy across projects.•Streamlined Reporting & Data Automation: Developed automated solutions for financial document data extraction and processing, using structured output for financial reports, ensuring efficiency and data accuracy.

India

•Ticketing Data Predictive Model: Conducted data analysis and developed an NLP-based predictive model using the LSTM algorithm to analyze ticketing text data, achieving 87% accuracy and reducing client manual efforts by 90%.•Receipt Data Extraction with NER: Applied NLP and Named Entity Recognition (NER) to extract structured information from receipts (Name, Location, Date, Total Amount), automating data handling and improving client data workflows.•Container Management Automation: Automated container startup and implemented continuous production monitoring, minimizing production downtime, and reducing manual efforts and operational costs for clients.•End-to-End ML Deployment: Deployed machine learning models end-to-end using Flask and FastAPI, ensuring scalability and smooth integration within production environments.

Bengluru, India

• NLP for Text Analysis: Utilized NLP techniques (NLTK and spaCy) for text analysis, enabling data comparison and improving text-based data insights.• Process Automation with Notifications: Developed automated workflows for redundant tasks, including email alerts using Exchangelib and mbox, significantly reducing manual workload and response time.• Data Extraction from PDFs: Automated data extraction from financial document images (Balance Sheets, Cash Flows, Income Statements) in PDFs using Tesseract OCR, delivering structured data to streamline reporting.Technical Skills:• Programming Languages: Python• Natural Language Processing (NLP): NLTK, spaCy, Named Entity Recognition (NER)• Optical Character Recognition (OCR): Tesseract for image-based data extraction• Automation Tools: Exchangelib, mbox for automated email notifications and workflows• Data Processing and Manipulation: Pandas, NumPy• Data Extraction and Parsing: Regex for text data extraction• Version Control: Git