Cristián Ormazábal
Software engineer building tools for complex systems.
How I Think About Engineering
I am a software engineer with over two decades of experience designing backend systems, developer tooling, and systems software.
My central focus is straightforward: I build tools for understanding, automating, and communicating complex software systems.
In practice, I tend to look for recurring patterns beneath individual problems, find the right abstraction, and turn that abstraction into deterministic tools, languages, or systems. Whether through structural graphs, declarative compilers, workflow engines, or database models, the goal is always finding the right abstraction for recurring complexity—giving developers clear mental models and repeatable leverage.
Recurring Themes
Codebase Understanding & Topology
Extracting structural ground truth from large codebases. Mapping symbols, call relationships, dependencies, and blast radius so that explicit structural knowledge complements text search and large context windows, helping engineers and agents reason about systems more directly.
Deterministic Visualization
Treating technical visualizations and data structure diagrams as declarative, version-controlled code. Compiling structured specifications into reproducible graphics that remain accurate and inspectable as systems evolve.
Executable Technical Communication
Transforming technical presentations, documentation, and demonstrations into executable artifacts. Creating tools that can navigate code, run commands, reveal diagrams progressively, and reproduce engineering workflows directly within the editor.
Domain-Specific Languages & Compilers
Designing concise grammars and focused compilers when general-purpose tools become too indirect or fragile. Encoding domain rules explicitly to enforce invariants and reduce recurring classes of mistakes.
What I Am Building & Exploring Now
My current engineering work sits at the intersection of developer tooling, software topology, declarative visualization, and AI-assisted programming.
I am exploring how to provide persistent structural knowledge for software agents and human engineers alike. As codebases grow, understanding systems requires queryable models of symbols, dependencies, call relationships, and runtime evidence. Grounding development workflows in an explicit structural model helps both people and autonomous agents reason about code without relying solely on heuristic searches.
Alongside topology, I continue refining declarative diagram compilers and programmable presentations that turn static documentation into executable workflows. Across all these efforts, the overarching goal remains the same: making complex engineering systems easier to inspect, explain, and operate.
Evidence in Practice
-
CodeTopo
A software-topology engine that builds a persistent structural graph of large codebases, combining static analysis and runtime evidence so humans and agents can reason about system structure directly.
-
Triton
A deterministic diagram compiler that turns declarative text into technical diagrams, data-structure visualizations, and composite technical layouts with live VS Code preview.
-
DeckPilot
A programmable presentation engine for VS Code that turns Markdown decks into executable technical presentations capable of navigating code, highlighting lines, running commands, and integrating live diagrams.