Senior .NET engineer from Chile, focused on enterprise APIs, integrations, automation, and applied AI. I turn unclear technical ideas into working systems that real teams can maintain.
I am currently exploring OpenAI Codex as a practical coding partner for software delivery: scoping tasks, implementing features, reviewing changes, improving tests, and documenting repeatable workflows.
- Enterprise backends and integrations with C#, .NET, ASP.NET Core, SQL Server and Azure.
- Applied AI workflows: RAG, agent-assisted development, local models, automation and evaluation loops.
- Full-stack prototypes with React, TypeScript, Next.js, Supabase, PostgreSQL and Docker.
- Developer workflows that combine human judgment with AI tools instead of treating AI as magic.
My main interest is not just generating code faster. It is learning how to use Codex responsibly in real software projects:
- breaking ambiguous work into verifiable tasks;
- keeping architecture simple enough to maintain;
- using tests, logs and reviews to validate AI-generated changes;
- documenting workflows so other developers can repeat them;
- sharing practical examples with the Chilean and LatAm developer community.
| Project | What it shows | Stack |
|---|---|---|
| pocia-count | Browser prototype for counting people and cars from a mobile camera feed. Built as a public Codex experiment. | React, TypeScript, TensorFlow.js, COCO-SSD |
| musicvideo | Spotify playlist browser with YouTube search integration. | Next.js, TypeScript, NextAuth, Spotify API |
| specboot | Portable rules and development standards for AI-assisted software projects. It is in WIP | AI coding workflows, project standards |
I am also building Las Cuentas Claras, a Chile-focused app for splitting and managing shared expenses in groups, trips, couples, families, and roommates.
The repository is private, but the project reflects the kind of work I enjoy: take a real everyday problem, design a simple product experience, connect web and mobile clients, and iterate with users instead of just building a technical demo.
Current scope includes:
- React/Vite web app deployed on Netlify;
- React Native mobile app in Android closed beta;
- Supabase backend with PostgreSQL, authentication, and serverless functions;
- payment and premium-flow experiments for a freemium model;
- product work around activation, retention and conversion.
You can find the live product at Las Cuentas Claras
I am moving deeper into Applied AI Engineering: combining my background in enterprise .NET systems with AI-assisted development, automation, RAG and product delivery. I also completed a diploma program in Generative AI and Language Models from CENIA, which helped me connect LLM concepts with practical software engineering decisions: architecture, evaluation, automation, privacy, and responsible use.
A good example of the work I want to keep sharing publicly is:
- choose a real engineering problem;
- build a small but functional solution;
- use Codex where it helps;
- verify the result with tests or manual checks;
- document the workflow clearly for other builders.
C# · .NET · ASP.NET Core · Azure · SQL Server · PostgreSQL · React · TypeScript · Next.js · Python · Docker · Supabase · n8n · OpenAI · Codex · RAG · Local AI
I am interested in helping more developers in Chile and LatAm use AI coding tools with better engineering practices: clear scope, version control, testing, reviews, security and maintainability.
Topics I like to share and discuss:
- practical Codex workflows for existing codebases;
- AI-assisted development without losing code ownership;
- .NET + AI for business applications;
- automation and integrations for small teams;
- lessons from building public prototypes.
- LinkedIn: linkedin.com/in/julioamestica