
Shubham
Data And AI Engineer
Habilidades

Conheça meus serviços


Experiência profissional
Amazon
1 yr
Business Intelligence Engineer
Sep 2025 - Mar 2026 • 6 mos
• Flagged likely B2B fraud for the Credit & Compliance team, as measured by a model scoring 3 signal types (synthetic business profiles, credit history length, email signals), by building a small machine-learning model. • Surfaced possible freight theft from vehicle sensor data, as measured by alerts on every unscheduled stop, by building a detection system over installed truck geo-location tracking. • Made every metric self-explanatory without needing to ask a person, as measured by 5 data points surfaced per metric (definition, refresh cadence, point of contact, architecture, ETL job), by independently building an AI tool to generate them. • Cut the time to trace a report back to its source logic, as measured by 1 auto-generated UML diagram per query replacing manual reverse-engineering, by having an internally built AI tool generate them directly from SQL. • Stood up the data foundation for the entire SWA program, as measured by 40 metrics served from one warehouse (14 Finance, 12 First Mile, 6 Middle Mile, 8 Last Mile), by designing it on a reusable semantic layer. • Sustained a change-success rate of 99 in every 100 across SWA releases, as measured by release stability under strict governance, by validating each data model against data quality checks before deployment.
Data Engineer Intern
Mar 2024 - Sep 2024 • 6 mos
• Got early, hands-on exposure to a not-yet-public AWS AI product, as measured by 1 year of combined Bedrock experience across two roles, by being one of the initial testers of Amazon Bedrock before public access. • Eliminated the manual data-processing bottleneck, as measured by 95 of every 100 manual interventions removed, by building automated scheduling, validation, and monitoring into the pipelines. • Cut end-to-end pipeline latency for 100+ downstream analysts, as measured by a 30 percent reduction in processing time across 1TB+ of data, by integrating 6 enterprise systems into a centralised data lake. • Made large-scale queries noticeably faster, as measured by a 15 percent cut in query latency, by optimising SQL and PySpark transformations across distributed environments.
Data Engineer (A.S.E)
Tata Consultancy Services
Aug 2021 - Sep 2022 • 1 yr 1 mo
• Measurably cut recurring production defects, as measured by a lower defect recurrence rate, by building modular Python data ingestion and SQL transformation pipelines with automated testing and Git version control. • Compressed defect turnaround from 3 days to same-day, as measured by throughput across 50+ Jira-tracked tickets, by investigating and resolving recurring data quality issues. • Shortened the path from code to production, as measured by CI/CD-driven deployments, by building Jenkins-based pipelines for automated testing and delivery.