Skills & Expertise (3)
Work Experience
Project Trainee
Centre for Artificial Intelligence and Robotics (CAIR), DRDO
10-2024 - 02-2025
Built a complete offline full-stack application using Python (Flask) and JustPy to run without internet connectivity, enabling secure deployment within the DRDO environment. Developed configurable text preprocessing pipelines, tokenization, stopword removal, regex-based cleaning, duplicate removal, stemming, and lemmatization, to convert noisy intelligence documents into structured, ML-ready text. Designed a dual-view interface for comparing the original document with the pre-processed output, improving clarity and usability for internal research teams. Implemented multiple word-embedding techniques such as Bag-of-Words, TF-IDF, and Word2Vec. Built a modular architecture allowing users to apply any combination of embedding methods and to transform documents into numerical vectors, enabling further pattern analysis and similarity detection Integrated clustering algorithms, K-Means, MiniBatch K-Means, Hierarchical Clustering, and DBSCAN, to group processed documents. Generated visualizations to help researchers categorize processed documents and assist research teams in identifying thematic patterns and hidden groupings. Delivered the solution as a fully offline, standalone Python system that meets DRDO s secure environment requirements and accelerates manual document analysis workflows.
Cloud Application Engineer Intern
Yoptima Media Solutions Pvt Ltd
07-2025 - Present
Improved the stability and visibility of the Client Data Platform (CDP) pipelines by adding error-handling, structured logging, and automated alerting. Set up contextual data-freshness checks in BigQuery that compare latest ingestion timestamps with expected T-1 values, reducing manual monitoring and helping teams quickly detect data delays and schema issues. Built an optimized alerting framework deployed on Cloud Run and triggered via Cloud Scheduler. Designed it to intelligently suppress duplicate alerts and notify only when data is newly delayed or recovered, enabling faster debugging for both technical and business teams. Designed and maintained BigQuery tables and scheduled queries powering the Superset dashboards. Ensured data consistency by integrating these tables with the alerting system for end-to-end monitoring of data updates. Optimized backend microservices and BigQuery SQL workflows, reducing data retrieval and preprocessing time by 30%, improving API latency for high-traffic endpoints supporting AdTech bidding pipelines. Wrote SQL transformations and mapping logic to unify Plan Line Item (PLI) data with third-party DV datasets, improving the accuracy and usability of performance dashboards for advertisers. Developed a lightweight YouTube metadata retrieval system for internal campaign optimization workflows. Built the frontend in React and backend in Python (Flask), deployed on Cloud Run. Added CSV validation checks and automated mapping logic to help business teams quickly fetch channel-level metadata via simple file uploads. Participated in enhancing CDP ingestion workflows by adding try/except logic, logging, and visibility alerts, improving reliability and reducing debugging time. Worked closely with Data Engineering and Business teams to understand reporting needs and support dashboard integrations, data validation, and backend improvements across multiple pipelines.
Education
B.E in Artificial Intelligence and Data Science
2023 - 2025 · India
Diploma, Civil Engineering
2018 - 2022 · India
Interested in this developer?
Profile Score Breakdown
Profile Overview
Availability Details
Visa Status
No Visa
Relocation
Open to Relocation
