• Built predictive models using logistic regression and decision trees to forecast customer performance and optimize retention strategies.
• Performed EDA on customer data and gathered insights from it to build an anomaly detection system.
• Evaluated performance of different LLMs like Meta Llama 3 8b and Meta Llama 3 70b based on their response to a prompt.
• Performed data wrangling and exploratory data analysis (EDA) on marketing datasets to identify trends and patterns influencing customer behavior.
• Applied unsupervised learning techniques like K-means clustering to segment customers into distinct groups, leading to targeted marketing campaigns.
• Collaborated with marketing and product teams to ensure data-driven decisions are made, enhancing the overall customer acquisition strategy.
• Worked on a POC for a Langchain and Vector DB based similarity search Generative AI tool.
• Developed Streamlit apps for marketing team and other stakeholders to visualize performance of the different existing and new machine learning models trained monthly by the team.
• Improve model training engine infrastructure by adding more robust unit tests.
• Researched vector databases to be used as a cache for a summarization tool.
• Orchestrated Jenkins pipeline by building custom configuration using Groovy.
• Collaborate with interdisciplinary teams to work on creating new tools and provide support for maintaining existing ones.
Skills: Machine Learning, Python, Vector database, Langchain, REST, Jenkins, SQL, EDA
