public class Engineer : IPolyglot, IArchitect, IAgentBuilder
{
public string Name => "Jorge Cura";
public string Philosophy => "Language Agnostic (The right tool for the job)";
public string Obsession => "Deterministic Systems & Context Engineering";
public string CurrentFocus => "Distributed Systems, AI Agent Orchestration & Developer Tooling";
public bool IsCoffeeDependent => true;
}Senior engineer with 7+ years building distributed systems that other engineers depend on. I've shipped high-volume payment infrastructure processing $50M+ annually, led production migrations without downtime, and recently focused on developer tooling and AI-assisted workflows.
I am language agnostic by design. Whether it's Rust for local-first performance, Python for agent orchestration, or C# for enterprise stability, I focus on building idempotent, fault-tolerant systems regardless of the stack.
I care about how systems interact, why builds break, and making the path to production shorter.
| Context | Technology |
|---|---|
| Languages | C#, .NET, T-SQL, TypeScript, Rust, Python |
| Backend | Microservices, RESTful APIs, Event-Driven Patterns, Idempotent Workflows |
| AI / Agents | Agent Orchestration, MCP (Model Context Protocol), LLM Integration, LangChain |
| Systems | Tauri v2 (Rust), Distributed Locking, Event Sourcing |
| Frontend | Vue.js 3 (Composition API), TypeScript, Pinia, Tailwind |
| Data | SQL Server, PostgreSQL, Redis, MongoDB |
| Cloud & DevOps | Azure, AWS, Docker, GitHub Actions, Terraform, CI/CD Pipelines |
| Testing | TDD, xUnit, NUnit, Jest, Playwright, Cypress |
Enterprise voice agents with real-time audio streaming.
- The Tech: Python, FastAPI, Twilio Media Streams, Google Gemini 2.5 Flash Native Audio, WebSockets.
- The Engineering: Architected a low-latency voice agent framework handling bidirectional audio streaming for natural, interruptible conversations. Built a stateful Orchestrator engine for multi-turn context, agent routing, and tool execution. Engineered an async "Intervention Observer" pattern that analyzes stream metadata for hotwords and negative sentiment in real-time without blocking the conversational loop.
LLM-powered customer intelligence β zero code, pure knowledge architecture.
- The Tech: 10-agent research team, 30+ academic sources (Bing Liu, Kahneman, Fitzpatrick, Kleppmann).
- The Engineering: Built a domain-specific AI skill with 27 analytical rules, 7-step methodology, RICE-scored recommendations, journey-stage mapping, competitive intelligence extraction, and cognitive bias detection. Demonstrates what AI-native product analysis looks like vs traditional NLP pipelines.
Bridging the gap between local filesystems and LLM reasoning.
- The Tech: Rust (Backend), Tauri v2, Vue 3, Tiktoken-rs, axum, SSE.
- The Engineering: Built a cross-platform desktop app for assembling AI prompts from codebases. Embedded an MCP server (JSON-RPC 2.0 over SSE) directly inside the Tauri binary β Claude Code connects to the running app and gets access to file scanning, prompt assembly, token counting, and clipboard through the same Rust engine the GUI uses. No external processes.
Cloud-native content delivery for the restaurant industry.
- The Tech: Cloud-first architecture, real-time device management, billing integration.
- The Engineering: Founding software engineer. Architected the platform from MVP to production β backend technology decisions, real-time device management, content delivery pipelines, and billing integration. Currently serving paying customers.
Security-first authentication and AI-driven engineering standards.
- The Tech: .NET 10, ASP.NET Core, WebAuthn/FIDO2, AI Agents.
- The Engineering: Implemented a passwordless authentication system, designed the public key infrastructure, and patched a critical security vulnerability. Built a multi-agent AI architecture to standardize engineering processes. Engineered a GitHub integration service with intelligent caching reducing external API calls by ~95%.
Building rails for millions in transactions.
- Resilience: Designed custom idempotency layers ensuring network retries never result in double-charges. Processing $50M+ annually with 99.99% uptime.
- Migration: Led zero-downtime migrations from legacy monoliths to modular Vue 3 micro-frontends serving 5,000+ daily active users.
I treat coffee brewing like a deployment pipeline: precise variables, consistent execution, constant debugging. James Hoffmann school of thought β maximizing extraction without hitting astringency.
# The Daily Driver (Aeropress)
config: { method: Standard, ratio: "11g : 200g", grind: Medium-Fine, temp: 100Β°C }
steps: [Pour 200g, Wait 2:00, Swirl, Press at 2:30 β stop at the hiss]If you want to talk about distributed systems, agent orchestration, or why your espresso tastes sour β hit me up.


