Alex Maeda

Senior Software Engineer with 9 years of experience building decision-support and workflow systems across ads industry, healthcare and logistics. Experienced in shipping AI-powered product experiences end to end - from React/Next.js interfaces to Python/FastAPI backend services, retrieval systems, and production ML integrations. A practical thinker who enjoys solving problems, collaborating with others, and always put the team's success first.

Software Engineer - Meta

(2024-12)

Shop Ads Team

• Launched an AI-powered onboarding experience for Shop Ads advertisers, developing React/TypeScript interfaces Python/FastAPI backend services that delivered Llama-based recommendations increasing feature discovery 10x by surfacing relevant tools contextually rather than through static navigation.
• Developed LLM-driven advertiser assistance workflows, integrating generative UI, GraphQL, and FastAPI-based orchestration services to deliver contextual recommendations during campaign setup replacing a static, menu-driven configuration flow and improving usability for new advertisers.
• Built and optimized FAISS-based vector retrieval infrastructure, improving indexing latency and backend efficiency to power AI recommendation features across Shop Ads workflows in a fast-paced ads platform.
• Reduced Llama inference latency by 45% through asynchronous processing, caching strategies, and database query optimization across Python/FastAPI backend services.
• Built monitoring and evaluation tooling for production LLM systems, supporting prompt versioning, response quality evaluation, and latency tracking across Shop Ads inference pipeline to support ongoing experiment and research.
• Shipped AI classification and anomaly detection services for financial integrity and ads monetization, identifying anomalous ad spend patterns and account behavior to reduce fraud risk and improve signal quality for monetization decisions.
• Standardized CI validation, continuous deployment observability, testing standards, and cross-team code review practices, improving engineering workflows across the Shop Ads AI team; mentored 2 junior engineers on LLMOps tooling, RAG architecture, and production AI system design.

Software Engineer - Sofar Ocean

(2022-01 - 2024-11)

Wayfinder Platform

• Built voyage planning product features within Wayfinder as part of a 6-person collaborative product team, developing React, TypeScript interfaces and Python/FastAPI backend services that integrated marine weather forecasts and route optimization models into decision-support tools for global shipping operators.
• Led frontend migration and refactor of legacy voyage monitoring workflows into a modern React component architecture with Jest test coverage translating ocean forecast and vessel performance data into clean, intuitive UX workflows that helped operators realize 4-8% fuel and emissions reductions per voyage.
• Engineered backend data pipelines and FastAPI services in Python that processed over 1.5 million daily ocean sensor observations, powering real-time voyage intelligence - weather routing, vessel performance, and fuel consumption insights across customer-facing APIs.
• Integrated route optimization logic into backend services, translating vessel performance model outputs predicting speed within 0.5 knots and fuel consumption within 2% accuracy into voyage recommendations balancing fuel efficiency, weather risk, and vessel constraints.
• Maintained the PostgreSQL data layer and AWS cloud services supporting forecast-driven voyage intelligence, including Docker workflows, CI/CD pipelines, and production monitoring for reliable model output delivery.

Software Engineer II - Uber

(2020-01 - 2022-01)

Uber Health

• Delivered Next.js dashboard features and Python, Django microservices for Uber Health's HIPAA-enabled transportation platform, supporting patient ride scheduling, trip monitoring, and coordinator operations across 3,000+ healthcare organizations.
• Architected REST APIs and distributed backend services for NEMT workflows including ride booking, real-time status tracking, and issue resolution reducing coordinator response time and improving operational visibility across patient transport workflows.
• Developed workflow automation and decision-support features for coordinator tools, building Django-based backend services that consumed and surfaced ML-powered predictions within Uber's broader Michelangelo ecosystem, streamlining patient ride assignment and reducing manual coordination overhead.
• Extended Uber Health's logistics platform with backend services and product logic supporting healthcare delivery workflows for prescriptions, groceries, and OTC items, connecting transportation and delivery operations for patients across multiple care settings.
• Strengthened deployment reliability and operational visibility across Uber Health's microservice architecture through CI/CD pipelines, production monitoring, and service-level observability, supporting full uptime for mission-critical healthcare transport services.

Software Engineer I - Uber

(2017-05 - 2019-12)

Uber Eats

• Developed backend services and APIs in Node.js and Python for Uber Eats' Restaurant Dashboard, supporting order intake, real-time delivery tracking, store operations, and menu management for 300,000+ restaurant partners across Uber Eats' global platform.
• Built and owned REST API design for restaurant order management workflows, handling real-time status updates, operational tooling, and issue resolution across merchant-facing systems serving millions of daily orders across 6,000+ cities worldwide.
• Integrated ML-powered outputs into restaurant operational workflows, surfacing personalized order recommendations and demand predictions that improved order conversion and partner retention across Uber Eats operations.
• Powered restaurant analytics and reporting features with backend services that gave merchant partners visibility into store performance, order trends, and business activity supporting data-driven decisions across a platform generating billions in annual GMV.

Software Engineering Intern - Google

(2016-05 - 2016-08)

Google Workspace

• Developed Python modules to integrate Google Classroom with Google Calendar for automated event scheduling within Google's EdTech Platform.
• Built responsive user interfaces using Angular, ensuring seamless integration with Google's design language and performance standards.

B.S. - Electrical and Computer Engineering - Carnegie Mellon University (2012 - 2016)