Gajraj Singh

Built Google ADK-based intelligent agents for marketing automation and conversational interactions. Designed and deployed scalable ML models on GCP, using services such as Cloud Run and Vertex AI, to support advertising and data pipeline automation. Led prototype-to-deployment work with a focus on cost efficiency and scalability, coordinating with distributed teams.

Developed an itinerary generator that validates inputs and produces customized travel plans using LLMs with LangChain. Contributed to a multi-bot suite, implementing a chatbot with LLMs, LangChain, and Pinecone for conversational data management. Worked on an education-focused app that generates bedtime stories with LLMs and processes text and image medical reports using Tesseract OCR to produce structured outputs.

Built a text-to-SQL prediction model for e-commerce use cases. Worked on machine learning tasks including a recommendation-oriented approach and an image similarity model using ResNet50-based cosine similarity. Used Python with PyTorch, scikit-learn, and pandas for data analysis and model development, and worked on an OCR-based ATR model for edtech platforms.

Handled database performance tuning and optimization to improve efficiency and reduce downtime. Implemented backup and recovery strategies to protect data integrity and minimize loss risk. Managed security measures and access controls, supported troubleshooting with cross-functional teams, and prepared client reports including performing MySQL migration.

Supported database migration projects across MySQL, IBM DB2, and Oracle environments. Monitored database capacity and usage, and investigated performance-related issues to prevent downtime. Contributed to operational tasks by identifying problems early and assisting in database maintenance activities.