Dippyoman Dutt

Ai Ml & Generative Ai Consultant @Wipro

Designed & led the development team of a Generative AI based chatbot, built a multi modal Retrieval Augmented Generation (RAG) system that can retrieve information from the text, images, tables present in the documents. Used various python libraries PyPDF2, PyMuPDF, Unstructured to parse the files. Researched and worked with different embedding models CLIP, all-MiniLM-L6-v2, text-embedding-ada-002 and with different vector databases ChromaDB, Faiss, Pinecone.2. Led the development team in building the Wipro GenAI QET Platform which uses Generative AI to automatically generate test scenarios, test cases, automation scripts, synthetic test data from user stories and business requirements. Developed the platform using Python, Streamlit, ReactJS, Node & LangChain. Integrated the platform with both enterprise version LLMs (OpenAI’s GPT models) and open source LLMs (Zephyr, Mistral & Llama). Built an AI ML solution using Python, Pandas, TextBlob, Matplotlib, Seaborn that scrapes google playstore, downloads the user reviews for the mobile android app of an UK banking client and applies NLP on the reviews to perform sentiment analysis. Built an AI powered chatbot using python, chatterbot library to respond back to customers of an UK banking client with the proper information and relevant links from their public website.5. Technically guided the teams in setting up automation frameworks, GIT related activities, DevOps, Orchestration, CI/CD. Developed automation scripts for an UK banking client’s mobile apps ( Android & IOS) using NodeJS, Javascript, Web driver IO, BDD Cucumber, XCUITest, Jasmine, Chai, Page Object Model (POM), Perfecto Cloud and Appium. Designed and worked on multiple automation frameworks in Lloyds for web applications using Java, Selenium Webdriver and Javascript, Web driver IO. Architectured API automation framework using Java, REST Assured.