Kyle Phillips

Founder & Lead Engineer

Nosdesk Pty Ltd

• Run the Nosdesk backend (Rust/Actix-web, PostgreSQL, Redis) as a containerised service on Fly.io, owning it end to end from architecture through production operation.
• Made deployments safe and hands-off: liveness and readiness probes, automatic schema migrations, and a CI gate for dependency vulnerabilities (cargo-deny).
• Built the real-time layer: an append-only Postgres change-event log pushed to browsers over Server-Sent Events via LISTEN/NOTIFY, so fan-out scales across instances with no message broker, plus Yjs CRDT collaborative editing over WebSockets backed by a Redis document cache and debounced snapshot persistence to Postgres.
• Enforced multi-tenant isolation with Postgres row-level security fronted by Rust's type system: handlers can't reach a raw connection, so cross-tenant access fails to compile rather than leaking at runtime.
• Instrumented the service with tracing, JSON structured logging with request-correlation IDs, and Grafana dashboards for throughput, capacity and latency; benchmark hot paths in development so inefficiencies are designed out rather than debugged in production.

Read the engineering deep dive →

Infrastructure Engineer

Solution One Pty Ltd

• Handled L2/L3 escalations across a multi-client MSP portfolio, working the root cause of problems spanning Linux servers, networking, web hosting, and EDR.
• Migrated the monitoring platform from Graylog to Grafana (Prometheus, Loki, Grafana Alloy, Alertmanager), with SNMP coverage of Cisco, MikroTik and UniFi devices and Windows security-event monitoring, so we caught incidents before clients reported them.
• Wrote custom Prometheus exporters in Go and Rust to instrument proprietary in-house applications and the OS resources around them, where no community exporter existed, and authored 10 alerting rule files that correlate metrics and logs to flag inconsistent behaviour before it becomes an incident.
• Kept the monitoring stack in Git and validated every change through an Azure DevOps pipeline (yamllint, docker compose, promtool) before deploying to Docker.
• Put network device configs under version control with Oxidized and Git across Cisco, FortiOS and RouterOS, and started building NetBox-driven config generation to move the team from manual changes toward automated configuration management.
• Wrote tooling in Python and Bash to migrate 150+ WordPress sites off two legacy cPanel servers onto CloudLinux 8, cutting over with minimal downtime and no data loss.
• Hardened Linux servers (SSH key and algorithm policy, iptables rate limiting) and automated the repetitive work: patching, SSL renewals across 20+ systems, and backup and EDR health checks, each written up as a runbook the rest of the team could follow.

Technical Support Specialist

Solution One Pty Ltd

• Owned networking and virtualisation alongside the IT Manager, running a segmented enterprise network on Palo Alto firewalls, stacked core switches, and Meraki wireless.
• Traced connectivity problems end to end, across 802.1X and SCEP certificate-based onboarding, VLANs, DNS, and DHCP.
• Set up self-hosted Outline docs and NetBox as the network source of truth, and wrote the runbooks and systems records the school still uses today.
• Brought Windows Server and VMware configs up to date, and migrated the helpdesk to Zammad, data migration and cutover included.

Founding Software Engineer

Neuroscale AI

• Short software engineering contract at an AI talent-intelligence platform: AI pipeline work, admin CMS features, and some security and architecture fixes across the wider system.