Sai Srinidhi Karanam

Data Science Analyst at Madhi.ai

Designed & deployed AI agent workflows using SLMs, Crew AI and advanced RAG to deliver context-aware, high-accuracy results. Built robust ETL pipelines to process PDFs, emails, CRM and other unstructured enterprise data, improving model performance and reliability. Created interactive dashboards in Looker to track KPIs, cost efficiency and adoption metrics, supporting data-driven decision making. Collaborated with engineers and product teams to fine-tune AI agents, reducing latency and increasing reliability in production.

Data Analyst at Armorblox

Developed automated data pipeline using Snorkel & keyword-based functions, improving labeling efficiency & model accuracy to 92%. Built and maintained dynamic, interactive Power BI and Python dashboards to visualize attack types, frequency patterns, and temporal trends, enabling data-driven decision-making, performance monitoring and strategic product enhancement prioritization. Analyzed and interpreted large-scale email security datasets using SQL, Python and advanced statistical modeling to detect fraud patterns and sensitive information, strengthening threat detection, data integrity and identifying zero-day phishing attacks. Researched, validated, and optimized multiple machine learning classification models to improve detection accuracy and reduced false positives by 12% through automated retraining, dataset updates, model evaluation and parameter tuning. Collaborated cross-functionally with threat research, product management and customer success teams to operationalize insights,