Matsu /ˈma.tsu/ — what friends call me.
Senior Software Engineer building with Agentic Engineering.
Modernizing the portal and backend infrastructure of an IoT platform serving millions of connected devices, evolving legacy systems and rebuilding core services to improve performance and reliability. Applying Agentic Engineering across software design, implementation, and delivery.
Re-architected legacy messaging systems into scalable microservices supporting over one million SMS messages daily, alongside WhatsApp and email integrations. Designed resilient AWS architectures for reliable, high-volume communication.
Contributed to CRM features and internal tooling, investigated customer-reported issues through application logs, automated internal workflows, and worked on WhatsApp chatbot integrations.
I use AI agents to explore, plan and implement software. The problem definition, architecture, technical decisions and whether the result actually works stay with me. This is how a change usually moves; small fixes skip straight to implementation, bigger work gets more exploration up front.
Frame the problem, the constraints and what done means. Architecture and trade-offs get decided here.
Agents read the code, surface options and draft a plan. I choose the direction.
Implementation goes to agents, scoped to the plan and the codebase’s conventions.
I read the diff, run the checks and test the behavior. Nothing ships without my sign-off.
Example · this siteEvery change starts as an issue. Agents follow CLAUDE.md, write an ADR when a decision matters and open a PR, which I review before it merges.
ExploringMulti-agent orchestration with Herdr and OMP. Early days, and still shaping how I split work between agents.