Ramesh G

• Demand forecasting using Time series forecasting, Supervised learning
• Created a automation script to evaluate similarity (Quality Analysis)
• Created a Python script with configuration file
• Applied Similarity Score between Human Created Solution for a Tickets and LLM generated solution for the tickets
• Fine Tuned LLM models (GPT, LLama, Mistral and other Transformer models) with customers and domain specific data to improvise the output quality
• Worked LLM & Generative AI techniques using Bert-Score, Bert, MicrosoftDeberta, Roberta and other models for Semantic Similarity tasks to enhance model performance
• Experience in LLM-based chatbot development, including integration and architecture patterns like RAG, Agent Assist architecture
• Stored the script in and config files in Github with version enabled and triggered using GCP Cloud Run
• Built end-to-end MLOps pipeline for Data Extraction, Pre-processing Big Data pipeliene, Model training, Prediction, Evaluation, Monitoring, Version control
• Created Data Pipelines from different data bases by integrating with automation scripts
• Deployed Data and ML Pipeline with Docker, Kubeflow and Vertex AI services
• Built recommendation engine to pick right solution and solve business problem
• Generated decision-making Data Visualization reports with multiple statistical values
• Build and deployed models using tools like GCP Vertex AI
• Help manager to build the division from ground up
• Experienced in Snowflake, SAP Enterprise HANA, Snowflake and Databricks
• Basic working Knowledge on Azure Data Factory
• Strong Skills in SAP HANA, SAP HANA procedures, script based calculation views, Snowflake, Databricks, pyspark, python
• Experienced in writing HANA SQL statements debugging and troubleshooting issues based in HANA DB
• Created Snowflake models by taking HANA models as a blue print
• Good oral, communication and presentation skills; Experienced with onsite-offshore coordination
• Enthusiastic, optimistic and good team player
• Strong team orientation and hard working

Machine Learning & AI Development:

· Developed classification, clustering, and recommendation models using XGBoost, Random Forest, Gradient Boosting, and NLP techniques.

· Applied Hyperparameter tuning and AutoML techniques to optimize model performance.

· Designed real-time fraud detection models using anomaly detection algorithms.

MLOps & Model Deployment:

· Developed CI/CD pipelines for automated model deployment using Git, GitLab, Airflow, and Terraform.

· Deployed AI models via Vertex AI, AWS Lambda, and FastAPI for scalable enterprise solutions.

· Implemented A/B testing frameworks for continuous model evaluation.

Data Engineering & Automation:

· Built ETL pipelines using BigQuery, SQL, Python, ensuring seamless data ingestion and transformation.

· Automated AI workflow scheduling using Apache Airflow to reduce manual intervention.

· Optimized database queries to reduce latency and improve data processing speeds.

AI Security & Monitoring:

· Conducted security audits and implemented SAST scans for compliance in AI deployments.

Integrated monitoring solutions for real-time model performance tracking and alerting.

AI Development & Model Optimization:

· Developed supply chain analytics models using time series forecasting and regression techniques.

· Built real-time customer segmentation models to enhance targeted marketing strategies.

· Applied transfer learning techniques for optimizing model training efficiency.

Predictive Analytics:

· Designed risk assessment models for fraud detection and customer churn prediction.

· Developed credit risk scoring systems using XGBoost, Scikit-learn, and Spark MLlib.

· Built real-time recommendation engines for e-commerce personalization.

Data Engineering:

· Implemented real-time data pipelines using Apache Kafka & SQL, reducing data processing time by 50%.

· Developed automated feature selection tools for improving model explainability.