Professional Experience

Architecting enterprise streaming solutions and leading AI/ML initiatives for major customers across AWS, GCP, and Azure. AI Subject Matter Expert focused on multi-agent systems, MCP integrations, and Flink-based AI architectures.

• Architected end-to-end streaming solutions for 50+ enterprise customers across AWS, GCP, and Azure, reducing ingestion latency and infrastructure costs
• Supported deals totaling over $15M+ in the past year as SE and AI SME
• Led AI/ML proof-of-concepts including anomaly detection, agentic workflows, MCP Servers, and A2A integrations
• Founded and host the Data+AI Monthly Internal Meetup, a recurring series for sharing knowledge on GenAI, multi-agent frameworks, and streaming AI architectures across teams
• Created and presented 'Every AI Opportunity is a DSP Opportunity', an internal/external deck on event-driven AI architecture patterns including RAG, MAS, and Flink Agents
• Authored technical content on 'CrewAI Crews as Apache Flink Jobs' and 'Flink Multi-Agent System Design', mapping multi-agent concepts to event-driven architectures
• Led technical architecture and PoC validation for high-profile LATAM accounts (Blip, iFood, Pipefy, Stix, Michelin, Superlogica) involving RAG, Flink ML inference, and platform modernization
• Collaborated with product and engineering to translate customer feedback into roadmap improvements, contributing to Confluent MCP, Flink Agents (FLIP-531), Streaming Agents, and Lightning Tables
• Delivered workshops and onboarding for engineering, data science, and DevOps teams on streaming design, partitioning, security (RBAC), and performance tuning
• Proactively secured early access to experimental GenAI features (vector search, model inference) for strategic LATAM customers, providing technical validation feedback to engineering

Led multiple engineering teams (15+ people) collaborating with directors and senior leadership. Cut AI inference latency by 20x and reduced costs by 6x.

• Managed multiple engineering teams (15+ people) with direct and indirect reports; mentoring, performance reviews, and strategic collaboration with peers at director level and senior leadership
• Improved RTA latency of AI inference pipeline by ~20x, enhancing user engagement on the online platform
• Drastically reduced unit costs (~6x) by rewriting ML components and optimizing K8s Clusters (GKE) utilization
• Redesigned core insurance product component, cutting new customer onboarding from 6+ weeks to hours
• Implemented full observability stack (SRE), reducing production incidents by 90% and achieving 99.9% platform uptime
• Implemented release process, full CI/CD automation, and observability strategy enabling 4 MLOps engineers to manage hundreds of production services processing millions of images daily
• Conceptualized, designed, and built MVP for innovative fraud detection platform (first of its kind) within one quarter, unlocking go-to-market strategies and revenue opportunities

Led globally distributed R&D team building reusable ML pipeline architectures for PwC's legal tech AI practice.

• Led team of 5 engineers across Argentina, India, and USA building reusable Data & ML pipeline blueprints for PwC Labs' legal domain projects
• Created Python SDK wrapping ABBYY FineReader Engine (C++ only), enabling ML scientists to prototype OCR solutions at 5x faster speed
• Architecture validated and adopted across 10+ legal domain projects before broader enterprise rollout

Built a fraud detection platform from scratch for the dental insurance industry, processing terabytes of dental images using AI.

• Hired, managed, and grew a globally distributed team of 12+ engineers (USA, Brazil, Philippines)
• Built AI pipeline processing terabytes of dental images, identifying a 0.5% fraud rate in the first release, enabling insurers to recover costs and take corrective action
• Designed the platform architecture from scratch to production, serving dental payers across the US

Led engineering team building search engine and marketplace product matching solutions. Created Marvin AI, the first Brazilian open-source project incubated by Apache Foundation.

• Managed team of 10+ engineers and data scientists creating components to improve search engine and product matching for marketplace
• Created Marvin AI, the first Brazilian open-source project incubated by Apache Foundation
• Received 'The Innovation Value Award' from PwC as most innovative project in 'Retail' category (2018)
• Delivered public presentations as speaker at international conferences on ML and Software Architecture (Apache conferences, QCon, PAPIs, SEI Saturn)
• Grew the open-source community and user base worldwide

Built an AI-powered e-commerce startup from scratch through acquisition by B2W Digital. Shipped 3 products using ML and data processing technologies.

• Hired, managed, and mentored a team of 20+ Software Engineers, DevOps, and ML Engineers
• Created and managed research partnership with Brazilian academic institution, resulting in published papers in computer vision and NLP
• Developed 3 products (B2B and B2C) using ML and data processing technologies as building blocks
• Built low-latency API for white-labeled image-based product search
• Successfully acquired by B2W Digital (2016)