B. E. student in AI & Data Science (CGPA: 8.29) with a software-engineering-first approach to ML: building, testing, and deploying ML/NLP services end-to-end rather than pure research. Comfortable with Python, SQL, and Git-based workflows, with hands-on experience designing evaluation pipelines to iteratively measure and improve model accuracy.
Shipped multiple full-stack ML systems (ingestion → model → evaluation → dashboard/API) independently, outside coursework. Looking to bring this build-test-deploy mindset to a production ML platform team while growing into cloud-native deployment (GCP) and CI/CD practices.
Cognify Technologies Pvt. Ltd. / Pune
Analyzed data trends using Python and SQL; wrote and debugged queries against production datasets to support business reporting Built Power BI dashboards for stakeholder-facing reporting, translating raw metrics into clear, actionable views Performed data cleaning, wrangling, and preprocessing on large raw datasets to prepare them for analysis and modeling – directly transferable to preparing data for production ML pipelines
B.E. in Artificial Intelligence & Data Science Expected
Savitribai Phule Pune University
Built an end-to-end ML/NLP pipeline applying text embeddings and a classification model to detect cybersecurity threats from log data in real time; designed an evaluation loop to iteratively test and validate model accuracy before deployment Designed a Mongo DB-backed ingestion/preprocessing pipeline (analogous to an ETL job) and an interactive Streamlit dashboard to monitor model performance and anomaly flags in a production-like setting Echo – Multi-Agent Customer Intelligence Pipeline Lang Chain, Lang Graph, Claude API, RAG, NLP Built a multi-agent pipeline that clusters customer feedback using NLP embeddings and predicts churn risk with a text classification model, including train/test evaluation to measure performance Implemented a self-validating critic agent that fact-checks outputs against source data – an automated evaluation layer – and generates structured reports; modularized agents into independent, testable components for easier debugging Edu Nav – Personalized Learning Path Platform Python, LLMs, Agentic AI, RAG, Prompt Engineering Designed LLM-powered agentic workflows with RAG to generate job-role-specific learning paths, grounding outputs in a curated knowledge base to reduce hallucination and improve reliability Applied prompt engineering to make responses deterministic and machine-parseable; built a conversational interviewprep module with iterative testing across prompts to refine response quality