Seraphine
Local multimodal assistant
PRIVATE PROJECTA personal interface combining local models, voice, a browser-based 3D avatar, persistent memory, and human-confirmed tool access.
SOFTWARE ENGINEERFULL-STACKDEVOPS
From the first interface to the infrastructure behind it all.
From artificial intelligence to embedded audio. A selection of what I build by connecting products, software, and infrastructure.
Local multimodal assistant
PRIVATE PROJECTA personal interface combining local models, voice, a browser-based 3D avatar, persistent memory, and human-confirmed tool access.
Distributed home audio
PRIVATE PROJECTA multi-room audio system built with ESP32 devices, Python services, real-time communication, and a control PWA.
Embedded audio streaming
PRIVATE PROJECTESP-IDF firmware that receives PCM audio from a home server over TCP and sends it through I²S, with persistent reconnection for embedded installation.
A SaaS product for architects where a photo of an invoice becomes classified items and updates the project budget, with human review and spending views by stage and supplier.
I’m Jonathan Farias — a software engineer, full-stack developer, and DevOps engineer. I work between the product people use and the infrastructure that keeps it alive.
I turn complex moving parts into clear, reliable experiences designed to evolve.
Clear interfaces, built for the people using them.
Connected APIs, data, and services.
Deployment, automation, and observability.
Software that connects to the physical world.
Interfaces connect to services and data. Devices integrate with the backend. Infrastructure supports delivery and operations across every layer.
Web interfaces, installable PWAs, and cross-platform apps.
Business logic, integrations, and persistence, with WebSocket for real-time communication.
Connected devices and audio streaming between software and hardware.
CI/CD pipelines, deployment, and observability support every layer.
A strong product also needs clear decisions, sustainable operations, and attention to what happens after deployment.
Understand the system. Before choosing a tool, understand the flow: people, data, risks, and constraints.
Build with evidence. Interfaces, logs, and metrics should make system behavior visible.
Operate with confidence. Delivery includes automation, observability, recovery, and evolution.
Have a project in mind or an opportunity to share? Let’s start a conversation.