AI Agent Delivery Platform
Architected and built the company's AI Agent Delivery Platform, shipping the full production stack and delivering the first live customer deployment in just three weeks.
I build products that help you ship faster and waste less time.
I'm Mario Guerra. For the last 18 years, I've built products, APIs, and AI systems.
At Microsoft, I led TypeSpec—an open-source API definition language. Today, nearly all Azure services use it to generate SDKs, cutting their development and review time by over 30%.
Currently, I focus on AI. I build products that automate the tedious parts of your work, so you can focus on solving real problems.
Where I've worked and what I shipped there.
IntelePeer/Aqurio
Own AI product strategy and the engineering execution behind it, from architecture through production rollout.
Architected and built the company's AI Agent Delivery Platform, shipping the full production stack and delivering the first live customer deployment in just three weeks.
Designed a healthcare patient-engagement platform on Temporal, Python, and PostgreSQL, with messaging workflows built for scalable, compliance-aware patient outreach.
An AI consultancy that deploys senior engineering agents to architect and launch products in days, not weeks. Helping startups and SMBs turn ideas into launch-ready tools with measurable ROI.
Created Orchid, a local-first proxy and visual debugger for AI pipelines. Captures API traffic at the transport layer to record LLM prompts, tool payloads, and decisions to a local SQLite database. Integrates with the Model Context Protocol (MCP) to expose trace queries to agents, enabling step-by-step visual debugging and deterministic offline replay testing.
Created SOLO (Single Operator Loop Orchestration), a framework that turns AI coding assistants into a production-grade engineering team of one. Slash-command workflows enforce spec-driven design, security-first architecture, and test-driven development, so every feature is architected, critiqued, and revised before a line of code is written.
Microsoft
Led product vision and strategy for TypeSpec, an open-source API definition language transforming API development. Guided Azure service teams as a member of the Azure API Stewardship Board, and defined the KPIs used to measure product success and iterate on real-world usage.
Led the zero-to-one launch of TypeSpec, delivering an API definition language that achieved 8-10x code reduction and 30% faster review times, delivering measurable developer productivity.
Prototyped a RAG-based solution to improve the model's responses, enhancing developer productivity through context-aware code suggestions with secure data handling.
Qualcomm
Led the development of tools and platforms for Qualcomm's Hexagon DSP, the core of Snapdragon modem and Neural Processing Unit technology. Served as de-facto product manager for Hexagon tools used by first-party software teams, improving on-time delivery by 35% through data-driven development processes.
Personal Projects
Self-directed work exploring where LLMs actually hold up in production.
Built RAG-enhanced chatbots using Graph API and Azure Foundry, leveraging LLMs to optimize workflows and demonstrate API-driven user support with secure data exchange.
Developed an LLM-based solution to summarize large documents, optimizing data pipelines for efficient, secure processing with a sliding window approach.