Research opportunities in Generative AI , with more than two years of experience in developing data science solutions for diverse clients. Expertise in designing and implementing Generative AI solutions, including prompt engineering and Retrieval-augmented Generation. Proven ability to analyze client data and provide actionable recommendations for integrating AI technologies into business strategies.
Analyzed and preprocessed structured and unstructured data related to healthcare product procurement. Extracted text from unstructured PDFs using Azure Document Intelligence, followed by preprocessing and data cleaning. Utilized Regex for extracting simpler fields and Generative AI (GenAI) for complex field extraction. Stored the structured data in a database for further analysis. Created visualizations to generate insights and enhance decision-making.
This project created models to recommend products based on size. As data updates daily, a model of two categories was created: one that generates vectors daily and retrains the model monthly. Based on the similarity of vectors and model output, recommendations were provided. Due to the huge amount of data, Databricks were used for model development, and MLOps for CI/CD. Classical machine learning was used to train models, and vector-generation models were used for vector creation. Airflow was utilized for orchestration, along with Databricks jobs.
The use case involves developing a question-answering system for a client whose data is stored in an Azure SQL database. The system’s output includes either raw data, visual representations such as graphs, or textual descriptions derived from analyzing the data, in response to user queries. Designed solutions by understanding the client’s requirements. Utilized OpenAI and different Azure services for the creation of the system.
The objective of this project is to develop a sophisticated autonomous chatbot capable of seamless communication with users, adept at managing interruptions and handling multiple intents concurrently. Integrated LangSmith to log each OpenAI call and perform effective prompt fine-tuning. Responsible for developing the complete flow of the chatbot and seamless integration of LangChain OpenAI function agents, along with deployment.
This chatbot helps to generate huge revenue through hotel bookings over a chatbot. Developed a conversational travel chatbot using the FastAPI framework to mimic human-like conversation and provide recommendations for hotels, places, and itineraries based on specified business rules. Used Redis Enterprise service to store prepared, cleaned data with their embeddings using LangChain and OpenAI. Deployed project on App Service with a CI/CD pipeline.