I'm a software engineer from Chile 🇨🇱 focused on building backend systems, AI-powered developer tools, and agentic workflows.
My background is primarily in Go, Python, TypeScript, cloud architecture, and distributed backend systems. More recently, my work has expanded into AI engineering, particularly coding agents, MCP, code intelligence, context engineering, retrieval strategies, and the evaluation of agentic systems.
I enjoy working at the intersection of software engineering and applied AI — turning LLM capabilities into systems that can actually understand repositories, use tools, execute workflows, and solve engineering problems.
I'm also founder & CTO at vocari.cl.
- 🤖 AI Engineering & Agentic Systems
- 🧩 Model Context Protocol (MCP)
- 🛠️ Coding Agents & AI Developer Tooling
- 🔎 Code Intelligence, ASTs & Repository Analysis
- 📚 RAG, Context Engineering & Retrieval
- 📊 LLM / Agent Evaluation, Observability & Token Usage
- ☁️ Cloud-native Backend Architecture
- 🚀 Go, Python & TypeScript services
I'm currently exploring and building around:
- AI agents for software engineering
- Large-scale repository understanding using ASTs, code graphs and semantic retrieval
- MCP servers for exposing tools and domain knowledge to AI agents
- Measuring coding agents beyond "it worked" — including tokens, cost, model usage, tool calls and outcomes
- Local and cloud LLM experimentation
- AI-assisted engineering workflows
- Enterprise AI adoption and developer enablement
- Open-source contributions to AI developer tooling
I believe the next generation of developer tools will require more than powerful models:
good context, good tools, measurable execution and strong software engineering.
I'm interested in how coding agents can move from simple code generation toward goal-oriented engineering systems capable of:
- Understanding large repositories
- Discovering relevant context
- Using external tools
- Planning changes
- Modifying code
- Running and validating tests
- Measuring the quality and cost of execution
I'm experimenting with approaches that combine:
AST → Code Graphs → Search → Retrieval → Agents
Tools and concepts I frequently explore include:
- AST / ast-grep
- Repository graph generation
- MCP
- CLI-based coding agents
- Structured code retrieval
- Semantic search
- Context optimization
- Multi-agent workflows
I'm particularly interested in measuring:
Task
↓
Agent
↓
Model + Context + Tools
↓
Execution
↓
Tests / Validation
↓
Tokens + Cost + Tool Calls + Outcome
Because evaluating an agent only by whether it eventually produced some code is not enough.
- Advanced agentic architectures
- Context engineering
- LLM evaluation and observability
- Agent harness design
- Code graph retrieval
- Multi-agent systems
- Local LLM inference
- AI-assisted software testing
I use open source both as a way to learn and as a laboratory for ideas around AI engineering.
Some of the areas where I'm contributing or experimenting include:
- Coding agent tooling
- Structured agent usage metrics
- MCP integrations
- Developer productivity tools
- Repository intelligence
- AI engineering experiments
My goal is not only to use AI tools, but to better understand how these systems work, how they consume context, how they use tools, and how we can evaluate them objectively.
🌎 Portfolio → jaimehernandez.dev 💼 LinkedIn → linkedin.com/in/devjaime 🐦 X / Twitter → @HsJhernandez 🚀 Vocari → vocari.cl 📬 Email → hernandez.hs@gmail.com
Build AI systems like software systems: observable, testable, measurable and maintainable.
I'm particularly interested in collaborating on projects involving AI agents, developer tooling, MCP, code intelligence, Go, Python and cloud-native systems.
- 🧠 Blog: UV, el gestor de paquetes ultra rápido en Python
- 🤖 LangFlow + MCP demo for fake news detection
- 🐍 Python + IA Workshops
