Metro Vancouver, BC · Cloud, Bare-Metal & Applied AI Engineering

Architecting Reliable Cloud, Resilient Systems & Ambient AI.

Senior Infrastructure Engineer specializing in Kubernetes platforms, declarative GitOps automation, zero-trust edge networks, and sub-second Applied AI voice integration.

Inspect Projects Interactive CLI (Easter Eggs inside)

ORCHESTRATION

K3s & AWS EKS

Automated GitOps deployments

NETWORKING

OPNsense & Mesh

WireGuard & Tailscale overlays

OBSERVABILITY

Prometheus / Loki

Distributed tracing & alerts

APPLIED AI

'Vi' Voice Agent

Sub-second Gemini edge pipeline

Capabilities

Core Competencies & Toolkit

Full-lifecycle engineering: from bare-metal virtualization & firewall policies up to distributed cloud clusters and intelligent agent runtimes.

Cloud & Orchestration

Zero-downtime containerized workloads, resilient control planes, and elastic multi-region architecture.

Kubernetes (k8s/k3s) AWS (EKS, IAM, VPC) Docker Traefik Helm

IaC & GitOps Automation

Reproducible, declarative codebases with automated linting, security scanning, and drift detection.

Terraform Ansible GitHub Actions ArgoCD GitLab CI

Bare-Metal & Edge Networking

Enterprise hypervisors, hardware passthrough, VLAN isolation, WireGuard site-to-site, and ZFS replication.

Proxmox VE OPNsense Tailscale TrueNAS / ZFS VLANs / BGP

Telemetry & Observability

Unified metric ingestion, actionable SLI/SLO dashboards, and proactive alerts before outages impact users.

Prometheus Grafana Loki / Promtail Uptime Kuma

Applied AI, Audio & LLM Orchestration

Engineering real-time ambient systems: sub-second prompt chaining with Gemini API, local model inference with Ollama/Whisper, microphone satellite streaming via Wyoming protocol, and Home Assistant automation.

Gemini 2.5 API Ollama / Local LLMs Whisper & Piper Python 3.12 Go Home Assistant OS MQTT

Deliverables

Featured Architectural Systems

In-depth technical case studies across physical silicon, cloud control planes, and applied intelligence.

APPLIED AI & VOICE SATELLITES • Sub-Second Processing

'Vi' — Distributed Edge Voice Agent

Ambient room intelligence system combining distributed microphone satellites with large language models. Features real-time voice streaming with sub-second wake-word detection, customized prompt chaining via Gemini API, fallback to local Ollama inference nodes during upstream network latency, and bidirectional MQTT automation with Home Assistant.

Python 3.12 Gemini 2.5 API Wyoming Protocol Whisper STT Home Assistant

vi-satellite.service

> wake_word: "Hey Vi" (99.2% conf)

> streaming PCM audio (16kHz)...

> stt: "adjust workspace ambiance"

> dispatching tool_call: ha.light_set

> response rendered in 390ms

Node: Jabra Speak 510 STATE: ONLINE

INFRASTRUCTURE & BARE-METAL • High Availability

Proxmox VE, K3s & Zero-Trust Mesh Infrastructure

Engineered a multi-node bare-metal hypervisor cluster powered by Proxmox VE and managed via Ansible GitOps automation. Segregated with an enterprise OPNsense firewall, dedicated IoT/trusted VLANs, WireGuard site-to-site VPN, and automated local/offsite encrypted ZFS pool synchronization.

K3s Kubernetes Proxmox VE OPNsense Tailscale Ansible ZFS

Telemetry & Topology

Firewall: OPNsense (Multi-VLAN)

Mesh VPN: Wireguard + Tailscale

Storage Sync: ZFS automated snapshots

Ingress: Traefik + Cloudflare TLS

Nodes Uptime: 99.98%

CLOUD & TELEMETRY • Production Reliability

GitOps Multi-Region AWS Infrastructure & Observability

End-to-end production automation using modular Terraform. Zero-trust VPC peering, multi-account AWS Organizations, automated Grafana dashboards with Prometheus alerts sent to Slack/PagerDuty, and CI/CD security linting via Trivy and Checkov.

Terraform AWS (EKS, S3, IAM) Prometheus Grafana GitHub Actions

Observed Outcomes

100% declarative code coverage
<10s detection on service anomalies
35%+ cloud spend waste reduction

Interactive Terminal Emulator

trofimenko@cli:~$

Use ↑ / ↓ for history · Click chips or type

zsh — trofimenko@can-west-01 (x86_64)

UTF-8

Session initialized: trofimenko.dev [node: vancouver-edge-01]
Type help to list all commands or try the easter egg chips above.

Background

Pragmatic Engineering, Zero Fluff.

Infrastructure shouldn't be fragile. Whether deploying multi-region Kubernetes clusters in AWS or designing hyper-reliable Proxmox nodes in a private homelab, the principles remain identical: declarative code, strict telemetry, zero-trust network boundaries, and automated self-healing.

With the explosion of applied AI, the edge between DevOps and intelligent systems has blurred. I work with both — configuring sub-second inference pipelines with the Gemini API, running edge voice satellites, and optimizing hardware compute nodes for reliable agent orchestration.

Based in Metro Vancouver, British Columbia (Pacific Time), collaborating with engineering teams and founders tackling serious infrastructure transformations.

LOCATION

Metro Vancouver, BC (PST)

CORE FOCUS

Platform, DevOps & AI Infra

STATUS

Active on High-Impact Projects

Daniil Trofimenko

trofimenko.dev

Platform, Cloud & AI Engineering

System / Kernel: Linux / Debian / Proxmox

Languages: Python, Bash, Go, HCL

Mesh Routing: Wireguard, Tailscale, BGP

Interactive CLI: Run neofetch →