Shilpa Dantuluri

• Lead Data Engineer within Advanced Analytics/Data Science, owning architecture, delivery and production support end-to-end for enterprise telecom workloads.
• Designed production data-pipeline architectures for various clients across the US, Europe and Asia regions, handling terabytes of data and billions of records with horizontal scalability and reliability.
• Partnered with product, engineering and operations stakeholders to translate ambiguous business problems into scalable data/AI solutions, leading discovery, architecture, rapid POCs, production implementation and operational hand-off.
• Re-architected a real-time text/image event pipeline for customer highlights and flashbacks, delivering 40% faster processing while improving scalability, flexibility and operating efficiency; replaced a third-party service and saved multi-million dollars annually.
• Designed and implemented LLM-backed services directly in the production highlights/flashbacks pipeline: personalized title generation from theme/image metadata via PySpark UDFs; multilingual translation into Spanish, Japanese, Portuguese, French and German; and vision-model image tagging.
• Built self-hosted multi-model platform using open-source LLMs on on-prem and AWS GPU infrastructure; selected hosted vs. self-hosted approaches based on workload quality, latency, cost, and data privacy.
• Building AI-driven self-healing Airflow operations on Kubernetes: locally hosted models triage DAG failures, distinguish infrastructure/resource failures from code/data errors, automatically rerun recoverable failures from the last successful step, and send root-cause context to on-call engineers.
• Rolled out AI agents for development and testing, including automated first-pass code review on every Git commit against codified standards for formatting, syntax, code structure and data-writing functions; expanded AI-assisted development and test generation across the team.
• Built an NLP pipeline that reduced analysis effort from days to minutes with higher accuracy, replaced a third-party vendor and saved approximately $1M annually, enabling customer-experience and marketing teams to act on top issues faster.
• Developed churn and customer-retention pipelines/models that reduced AT&T zero-usage customers by approximately 20% and Virgin Mobile Latin America churn by approximately 13%.
• Contributed to applied marketing intelligence for offer optimization and response-rate optimization, supporting campaign lifts of 175% for telemarketing and 200% for direct mail.
• Built real-time IoT pipelines and dashboards tracking device runtime, energy use, temperature and humidity, contributing to approximately 3% energy savings for Rackspace and 5% for Yumi Ice Creams.
• Led migration of production applications from Hadoop-MapR to Kubernetes/AWS/Airflow, improving scalability and reducing operating cost.
• Created enriched datasets, scoring logic, and fact/dimension tables for millions of customers, enabling application UIs and Tableau dashboards for campaign optimization.

Environment: Python, PySpark, SQL, Spark, Hadoop/MapR, Kubernetes, Docker, Airflow, AWS, LLMs/GenAI, Prompt Engineering, GPU Inference, Git/Jira

• Analyzed large volumes of real-time patient, member and claims data across ODS, Datamart, CareEngine and Active Advice using complex SQL/PLSQL to isolate data issues, dependencies and business impact.
• Performed deep-dive analysis on recurring production inquiries; identified upstream/downstream patterns and designed validation and reconciliation scenarios that closed source-to-target gaps and reduced repeat incidents.
• Conducted deep-dive analysis on recurring production inquiries, identifying upstream/downstream patterns and designing validation and reconciliation scenarios that closed source-to-target gaps.
• Optimized SQL and PL/SQL stored procedures/functions while monitoring high-volume batch/reporting jobs to ensure performance and reliability.
• Developed and automated business/customer insight reporting using SQL Server, Oracle, Power BI, and SSIS; standardized recurring requests into repeatable reports and contributed to analytics warehouse/ODS design and data quality/performance testing.

Environment: Oracle, SQL Server, Informatica, SQL, PL/SQL, UNIX, Power BI, SSIS, Quality Center, ServiceNow, BMC Remedy, TFS

• Automated management reporting with Excel/VBScript and SSRS, scheduling daily, weekly, monthly, and quarterly reports for senior management, streamlining reporting processes.
• Environment: SSRS, Excel/VBScript, SQL/Reporting, enterprise metrics
• Self-motivated, with a strong sense of personal responsibility.
• Worked effectively in fast-paced environments.
• Skilled at working independently and collaboratively in a team environment.

• Analyzed large claims and overpayment datasets using SQL across Oracle, DB2, and SQL Server to identify overpaid claims, recovery candidates, and data-quality issues, enhancing data accuracy and recovery strategies.
• Traced claims lineage across legacy sources and target warehouse; validated Informatica mappings/transformations and reconciled record counts and financial totals, ensuring data integrity and compliance.
• Supported ETL testing/UAT/production validation, including claims test datasets and flat-file testing, ensuring accurate ETL monitoring and comprehensive translation of requirements to mappings.

Environment: Informatica 8.x, SQL Server 2008, DB2, Oracle log, SQL, PL/SQL, HP Quality Center, IBM Rational ClearCase/ClearQuest

• Developed Member Individualization web application and Clerical Tool search module to enhance user experience and streamline member data retrieval.

Environment: IBM Mainframes – MVS/OS390, DB2, COBOL, JCL, CICS, REXX, VSAM, TSO/ISPF