Software Engineer | Curious by Nature · Always Learning, Exploring & Building | AI · Cloud · Full Stack

Hi, I’m Iván
I love building things

I’ve spent nearly 20 years learning, experimenting and building software — from backend platforms and cloud infrastructure to open-source tools and applied AI. I’m happiest when I’m exploring a new idea, writing code and making something genuinely useful.

Portrait of Iván Corrales
Now building

Projects with purpose

Some projects are social initiatives, others experiments to explore how new technology works. All are part of a hands-on approach to engineering.

Social impact · 2026 — present

Leonoa Lab — Accessible technology

Building a non-profit initiative around free, accessible software. Its Adaptive Picto Communicator explores augmentative and alternative communication through pictograms, customizable vocabularies and user profiles, with ARASAAC integration.

Leonoa Lab ↗
Under active development

Working across the API, data model, automated tests and an end-to-end prototype. The project is evolving; it is not presented here as a finished product.

PythonFastAPIPostgreSQLSQLAlchemypytestAccessibility
Community & education · 2025–2026

Tech HSpes / FP Lumara

Helped build and coordinate a real engineering environment for students developing inclusive software. Teams used GitLab CI/CD, Kubernetes, Helm, Spring, NestJS, React, Flutter, Redis and RabbitMQ, while learning through delivery and collaboration.

Projects were recognized with a Premio Destacando. The initiative was supported by FP Lumara and Fundación Humana Spes.

Project announcement ↗
KubernetesGitLabJavaReactFlutter
Personal project · 2026 — present

AI Movie Studio

An experimental local-first creative platform for stories, images and conversational workflows. Exploring agent orchestration and local LLMs alongside an API and persistence layer.

PythonFastAPILangGraphOllamaPostgreSQL
Applied AI · 2026

Time-aware RAG with earthquake data

A reproducible experiment with real USGS data, historical versions, retrieval, deterministic workflow routing and local models. Built to answer not only what is known now, but what was known at a specific moment.

Read the case study ↗
LangGraphQdrantOllamaPostgreSQL
Engineering experiment · 2026

FieldOps — Offline-first Android

A small Android maintenance app used to investigate the difference between storing local state and persisting synchronization intent, including failure injection and measurable local-first behavior.

Source code ↗
KotlinComposeRoomWorkManager
Made to share · Open source

Tools I built for other builders

Before my current AI and social-impact work, I spent years building and sharing developer libraries. Some found an audience well beyond my own projects. These are selected original projects, not a list of forks.

Go · Functional programming · 2018500+ GitHub stars*

Koazee

A stream-like library for working with Go slices, exploring functional transformations, immutable operations and lazy evaluation. One of my earlier open-source projects to gain traction in the Go community.

Elixir · Testing · 2016Published on Hex

Exkorpion

A behavior-driven testing framework for Elixir, extending ExUnit with readable scenarios and Given–When–Then steps. Published on Hex and featured in community-curated Elixir resources.

Go · Library

Knoa

A utility for manipulating unstructured data, navigating nested values and converting between formats.

View source ↗
Go · Generics

Dataz

Reusable utilities for Go collections and data structures, taking advantage of generics.

View source ↗
Go · Aspect-oriented programming

Beyond

An exploration of aspect-oriented programming patterns in Go, published under Wesovi Labs.

View source ↗
Test automation

Orion

A testing tool with a domain-specific language for authoring and automating acceptance tests.

View source ↗
Go · API tooling

REST & Test Generator

An experiment in generating API service code and CRUD tests from a structured specification.

View source ↗
MCP · Observability

Observable MCP Server

A practical project exploring logs, traces and metrics for MCP-based systems.

View source ↗

*Koazee's star count is presented as an approximate historical/community metric, not a live counter. Several projects live under Wesovi Labs rather than my personal GitHub account. Older experiments may not be actively maintained.

Selected career

Built across the stack

Hands-on engineering across backends, data, frontend, infrastructure, architecture and product teams.

2025–2026

Tech Humana Spes / FP Lumara

IT coordinator & software development educator. Built the center's engineering infrastructure and CI/CD, coordinated DAM, ASIR and SMR, and mentored students building real projects.

2024–2025

Seen · via Wesovi Labs

Senior software engineering for an international product team: Go backend, React interfaces, FFmpeg-based video processing and Google Cloud deployment.

2023

AstroKube

Head of Engineering, designing and implementing core Go components for a multi-cloud Kubernetes operator SaaS, with Helm, Argo CD and GitOps.

2022

Smallpdf

Staff engineering on payments and purchasing flows with Go backend libraries, REST APIs, Adyen, React and TypeScript.

2021

Greenpeace International

Data engineering pipelines for ingestion, transformation and synchronization using Python, SQL, Airflow, Airbyte and Google Cloud.

2017–2020

BBVA

Software architect, squad lead and cloud tech lead, contributing to a multi-product internal cloud platform through APIs, asynchronous services and infrastructure automation.

Earlier work includes independent consulting for financial services, NewBay Synchronoss, Ibermática / ONCE, Kotasoft and Universidad Complutense de Madrid.

Notes from the field

Writing & experiments

All individually identified articles from Notion, Medium and LinkedIn, including the early Elixir, Go and machine-learning series.

From a Functional MCP Server to an Observable System ↗

Tracing, logs and metrics across an observable MCP server.

2026 · Notion · AI
Guardrails: cómo construir agentes más seguros y fiables ↗

Patterns to make AI agents safer, more reliable and easier to evaluate.

2026 · Notion · AI
Data Changes: How I Built a Time-Aware RAG ↗

Temporal retrieval with historical data and local models.

2026 · Notion · AI
Room doesn’t make you offline-first ↗

Why local storage alone is not enough for robust mobile sync.

2026 · LinkedIn · Testing
Getting Started with OpenTelemetry in distributed Go Microservices ↗

Instrumenting Go services with traces, metrics and logs.

2024 · Medium · Cloud
WebAssembly: Running Go code in the browser ↗

Compiling Go for the browser and understanding the runtime model.

2022 · Medium · Cloud
Build your own DSL with Go & HCL ↗

Designing a custom language layer on top of Go and HCL.

2022 · Medium · Cloud
BERT: Understanding the Null response threshold in QA systems ↗

A closer look at confidence thresholds in question answering.

2022 · Medium · AI
Transformers for Natural Language Processing ↗

An introduction to transformer models and NLP concepts.

2021 · Medium · AI
Minikube on a VPS from scratch ↗

Bootstrapping a compact Kubernetes environment on a VPS.

2022 · Medium · Cloud
Choosing the right programming language ↗

Trade-offs when selecting the right language for a project.

2022 · Medium · Cloud
Orion in Action: Testing a Real API ↗

Using Orion against a realistic API workflow.

2021 · Medium · Testing
Orion: A next-generation automation testing tool ↗

The ideas behind an automation testing tool for modern teams.

2021 · Medium · Testing
3 days playing with Rust (and I love it!) ↗

Early impressions while learning Rust by building and exploring.

2020 · Medium · Cloud
Go performance tips ↗

Practical tips to write faster, leaner Go applications.

2018 · LinkedIn · Cloud
Detecting bullying with a multilayer perceptron

Neural networks applied to a bullying-detection problem.

2017 · LinkedIn · AIDirect article URL not verified
Linear regression with Tensorflow

Exploring linear regression and the relationship between advertising spend and revenue.

2016 · LinkedIn · AIDirect article URL not verified
Testing your Maru Rest Api’s with Exkorpion ↗

Testing Maru APIs with an Elixir BDD library.

2016 · LinkedIn · Testing
Pass through authentication in Openldap ↗

Delegated authentication using OpenLDAP, SASL and Active Directory.

2016 · LinkedIn · Cloud
Implementing a REST Api with Maru (Elixir) ↗

Building REST endpoints with the Maru framework in Elixir.

2016 · LinkedIn · Cloud
Getting started with Elixir - Ecto

An introduction to the Ecto database toolkit in Elixir.

2016 · LinkedIn · CloudDirect article URL not verified
Implementing a pairs matching game in Go

Learning Go through a small pairs-matching game.

2016 · LinkedIn · CloudDirect article URL not verified
First steps with Erlang ↗

A hands-on introduction to Erlang and functional programming.

2016 · LinkedIn · Cloud
About

Curious across disciplines... Grounded in code

I have worked as an engineer, architect and technical leader, but what keeps me motivated is understanding problems and building useful software. My experience spans distributed backend systems, platforms and DevOps, frontend development, data engineering and applied AI.

I enjoy learning by experimenting, writing about the details and sharing what works — and what doesn't.

Toolbox, not a checklist

Backend & systems

Go · Java · Python · Node · Spring Boot · FastAPI · REST · GraphQL · gRPC · Event-driven systems

Platforms & data

AWS · Google Cloud · Kubernetes · Docker · Terraform · Helm · CI/CD · PostgreSQL · MongoDB · Redis · Kafka

Frontends & applied AI

React · TypeScript · Ollama · LangGraph · Local LLMs · RAG

Education

Computer Systems Engineering (UCM) · MSc in Artificial Intelligence (UNIR) · Master's in Teacher Education (UNIR)

Let's connect

Have an interesting problem to solve?

I'm particularly interested in thoughtful software engineering, open source, applied AI and technology with positive social impact.

Write directly to ivan.corrales.solera@gmail.com