Ivan Acosta-Rubio

// AI Engineer | Testing & Automation Specialist

// 📍 Miami, FL | US Citizen (No sponsorship required) | Remote (US Eastern)

// 💼 Available Immediately for Full-Time Roles | Open to Contract & Consulting

📅 Schedule a Free 30-Min Call | 📄 Resume PDF GitHub Twitter

Summary

Self-taught AI engineer with 17+ years building scalable systems and automation platforms. Specialized in building AI-driven testing infrastructure and developer productivity tools. Created AI systems that auto-generate tests from specifications and videos, reducing manual QA effort by 70% while scaling testing coverage across mobile and API platforms.

Technical Skills

AI/ML Engineering: LLMs (OpenAI, Anthropic), Context Engineering, RAG Systems, AI-driven Test Generation, Prompt Engineering, Model Integration

Programming Languages: TypeScript, Rust, Ruby, Python, JavaScript, Swift

Testing & Quality: Maestro, E2E Testing, API Testing, Mobile Testing (iOS/Android), Test Automation, TDD, Integration Testing, Load Testing

Cloud & Infrastructure: AWS, Docker, CI/CD Pipelines, GitHub Actions, Deno Deploy, Serverless Architecture

Databases & Storage: PostgreSQL, Redis, SQLite, Database Design, Query Optimization

Tools & Platforms: Deno, Bun, Node.js, OpenAPI, CLI Development, Git, REST APIs, GraphQL

Architecture & Design: Microservices, SDK Design, API Design, Facade Pattern, Event-Driven Architecture, System Design

Leadership & Soft Skills: Team Building (2-35 engineers), Technical Mentoring, Cross-functional Collaboration, Engineering Process Design, Stakeholder Management

Methodologies: Agile/Scrum, DevOps, Continuous Integration, Test-Driven Development, Remote-First Teams

Experience

Strike / Lead Test Engineer

2021 - October 2025 | Remote

  • Built Testing department from zero to supporting 35 engineers across Platform, Trust & Safety, Consumer, and Data teams
  • Architected and built SITS (Strike Integrated Testing Strategy), a full-stack AI-powered testing platform using TypeScript/Deno with RAG (Retrieval-Augmented Generation) system that ingests PRDs, design documents, and API specs to auto-generate comprehensive test plans and executable test suites
    • Implemented context-aware prompt engineering with Anthropic Claude and OpenAI GPT-4 to understand product requirements and generate domain-specific test scenarios
    • Built document parsing pipeline that extracts structured data from Markdown, Confluence, and Notion docs using custom NLP preprocessing
    • Created PostgreSQL-backed test case repository with version control, enabling teams to track test evolution across product iterations
    • Integrated with GitHub Actions CI/CD to automatically validate PRD completeness and suggest missing test coverage before development starts
    • Reduced test creation time by 65% (from 8 hours to 2.8 hours per feature) while improving test quality and coverage consistency across teams
  • Engineered video-to-Maestro AI agent that converts screen recordings into production-ready E2E test automation using computer vision and LLM orchestration
    • Built multi-modal AI pipeline combining GPT-4 Vision API for frame analysis with custom TypeScript orchestration layer to identify UI elements, user interactions, and app state changes
    • Developed intelligent YAML generation engine that translates visual interactions (taps, swipes, text input) into valid Maestro test syntax with assertions and wait conditions
    • Implemented frame differencing algorithm using image processing to detect meaningful UI changes and filter out redundant frames, improving processing speed by 3x
    • Created CLI tool with Deno that integrates with local development workflows, allowing QA engineers to record-once and generate multiple test variants
    • Deployed as serverless function on AWS Lambda with S3 video storage, processing 50+ videos weekly across iOS/Android platforms
    • Eliminated 30+ hours of manual test writing per week while achieving 95% accuracy in test generation, enabling QA team to focus on exploratory testing
  • Built CLI tool that auto-generates SDK from OpenAPI specs and provides testing facades, enabling self-service testing for 40+ developers
  • Designed E2E infrastructure for iOS/Android using Maestro, achieving 78% test coverage and reducing regression testing time from 4 days to 6 hours
  • Reduced critical API testing time by 80% through automated test generation matching real mobile app behavior

ZR Per Diem Services / Principal Software Engineer

February 2017 - 2021 | NY

  • Architected mission-critical application processing $10M+ in annual revenue
  • Integrated 12+ third-party APIs and built data processing pipelines handling 15,000+ transactions daily
  • Designed fault-tolerant systems achieving 99.7% uptime for business-critical operations

SoftwareCriollo / Founder & Staff Software Engineer

May 2008 - January 2017 | Miami, FL

  • Founded and scaled remote-first consultancy from 2 to 23 engineers across multiple client engagements
  • Led development of 35+ full-stack web applications as Staff Software Engineer, architecting backend systems, databases, and deployment pipelines
  • Built production Ruby on Rails and TypeScript applications serving 500K+ users with microservices architecture
  • Designed and implemented PostgreSQL and Redis database systems with query optimization achieving <100ms response times
  • Created comprehensive CI/CD pipelines using GitHub Actions, Jenkins, and Travis CI, reducing deployment time from hours to minutes
  • Implemented DevOps infrastructure including containerization with Docker, automated testing, and monitoring solutions
  • Managed cloud infrastructure on AWS (EC2, RDS, S3, Lambda) handling millions of API requests monthly
  • Led ops-heavy initiatives including database migrations, zero-downtime deployments, and incident response procedures
  • Released open source libraries including Ruby core contribution used by millions of developers

Education & Continuous Learning

Self-Taught Engineer with 17+ years of hands-on experience building production systems

Continuous Learning & Certifications

  • University of Washington - Machine Learning Foundation (2016)
  • Active contributor to open source communities and Ruby core
  • Speaker at international tech conferences (RubyConf Brasil, Inspect SF)
  • Published technical author on iOS, Swift, and AVFoundation

Formal Education

  • Miami Dade College - Associate of Arts (2005-2007)

Open Source & Notable Projects

Ruby Core Contribution - "Did you mean?"

Created misspell ruby module that was integrated into Ruby core as the "Did you mean?" error feature, now used by millions of Ruby developers worldwide. Implemented using Levenshtein distance algorithm.

Talks & Publications

Kiteboarding

Train alongside world champions. When I'm not coding AI systems, I ❤️ to kiteboard.

Contact

Consulting Availability

// While actively seeking a full-time role, I'm available for consulting to stay hands-on and help companies with their AI initiatives.

Available for AI engineering and testing automation consulting engagements.

Engagement Options

→ Free 30-Minute Consultation

Discuss your project needs and explore how I can help.

→ Paid Engagements

  • 1 Hour: $420 - Deep-dive technical consultation
  • 1 Week: $21,000 - Sprint-based implementation or audit
  • Monthly Retainer: $42,000/month - Ongoing advisory and development

// All engagements include technical architecture, code review, and implementation guidance