Daniel Porrello

Architected metadata-driven ETL/ELT pipelines on AWS and GCP using Python microservices, Spark, and event-driven orchestration, cutting p95 data refresh time from 45 minutes to 8 minutes and enabling intelligent schema discovery. Integrated LLM capabilities via Python APIs and LangChain-style orchestration to support natural-language data analysis and metadata enrichment while maintaining secure access controls, improving dashboard adoption by 35%. Built backend microservices and CI/CD using Docker and Terraform (with Azure) to improve secure deployments, observability, scalability, and incident recovery. Led technical standards across data, security, and platform teams, aligning metadata management and knowledge-graph/automation efforts with business objectives.

Built scalable Python backend APIs and distributed systems for real-time vehicle data analytics using Kafka, Azure, and caching, improving query performance by 40% and sustaining 99.9% uptime while supporting 120,000 daily events. Streamlined telematics ingestion automation by combining Airflow, Python, Docker, and LLM-powered classification, reducing manual review by 70% across 15 data sources and 6 squads. Optimized data integration across microservices using GCP and Spark lakehouse patterns, strengthening metadata management, star-schema generation, and security controls; reduced duplicate ingestion by 55% and analytics defects by 30% via automated validation pipelines.

Engineered Python microservices and data modeling workflows on Google Cloud using dbt and Airflow, improving query performance by 38%, data quality by 25%, and supporting 50M rows across multiple business objectives. Developed LLM-based semantic discovery using orchestration and vector search, reducing mapping time by 60% across enterprise domains while improving accuracy across squads. Established MLOps/LLMOps pipelines for responsible AI with observability, testing, and cost-aware scaling on AWS/Azure/GCP, increasing deployment frequency and reducing defects. Championed backend engineering best practices across teams, improving delivery and reliability.