Applied AI & Multi-Agent Systems | Flutter & Native Mobile | Full-Stack Architecture
Building intelligent, agentic products end-to-end
I'm a full-stack software engineer and architect with 10+ years shipping production mobile and cloud systems — now focused on applied, agentic AI.
I specialize in mobile — Flutter (iOS/Android/web/desktop) and native iOS (Swift, SwiftUI, UIKit) — and build full-stack around it: Node.js services, cloud architecture on AWS and Azure, and real-time IoT. Today my biggest focus is AI: multi-agent orchestration, MCP integrations, and RAG pipelines that turn frontier models into reliable production software.
I've delivered 14+ apps for 4+ Fortune 500 companies (Pinterest, Netflix, Intel, United Airlines) at a scale of 400M+ users. As a technical leader I own architecture end to end, mentor teams, and set the standards — most recently migrating a core product to an AI-driven, multi-agent architecture.
Designing agentic systems and shipping AI into production — from multi-agent orchestration to cost-optimized inference.
Personal Autonomous AI Assistant
A self-built agentic platform for prototyping multi-agent orchestration, RAG, RLM (Recursive Language Models), and context engineering across local & cloud LLMs — with MCP tool-calling against frontier models like Claude and Gemini.
Coordinating specialized agents with planning, delegation, and shared context.
Building MCP servers & tool integrations that give models real-world reach.
Retrieval pipelines and semantic search backed by QDrant.
Structured prompting and context windows tuned for accuracy and reliability.
Deterministic guardrails plus inference tuning for reliable, efficient models.
~30% inference cost cutEval harnesses, tracing, and metrics that keep model behavior measurable.
Automated pipelines for shipping and promoting models to production.
Personal projects I build in the open — code you can read, run, and review.
Multimodal RAG chatbot that answers only from your documents — every answer cites the file, the page, and an image of the source page. Python, Gemini embeddings, and Qdrant vector search, fully Dockerized.
Pokédex-style iOS discovery app built with SwiftUI, Clean Architecture + MVVM, and an offline-first SwiftData cache — tested across every layer and live on the App Store.
Procedural chiptune engine that composes seamless, loopable game music from a mood and a seed — deterministic, byte-reproducible output in pure Python with zero dependencies.
Open-source AI code-analysis tool by zendizmo that uses Recursive Language Models (RLM) to analyze codebases beyond normal context windows
My contribution: added full Flutter/Dart support — parsing, framework detection, mobile-aware prompts — and fixed 3 correctness bugs, submitted upstream as an open PR
Major apps and platforms I've contributed to
Led the migration of this product's core to an AI-driven, multi-agent architecture — agent orchestration, MCP integrations, and LLM-powered features in a Flutter app
Developed comprehensive healthcare platform for autism services, connecting clients, therapists, and families
Built companion app for autism therapy providers with AI-powered insights for personalized treatment plans
Built revolutionary wearable platform measuring nitric oxide and muscle oxygenation in real-time with AI coaching
Contributed to iOS development for one of the world's largest visual discovery platforms with over 400M users
Built internal mobile application enabling Netflix engineers to test and certify 1000+ TV models
Developed comprehensive campus management platform for Intel employees with AI-powered heat maps
Developed features for United Airlines mobile app used by millions of travelers worldwide
Created viral COVID-19 vehicle restriction app that reached #2 in Apple Store News category in Costa Rica
Delivering production software for Fortune 500 teams and high-growth products.
All applications I've developed and contributed to throughout my career
I'm always interested in discussing new opportunities and challenging projects.