Local inference
RTX A2000 12GB passed through to the GPU VM. Ollama serves qwen3.5:9b, qwen3:8b-16k and nomic-embed-text; Whisper runs speech-to-text on CUDA; Open WebUI is the front door.
matheesan.ca / lab
ch.03 / labWhere the models meet the metal. Production agent loops and eval tooling up top; local inference, computer vision, storage and a slightly overbuilt network underneath.
ml lab
Production systems taught me that intelligence is the demo; reliability is the job. The loop below is the part I actually build: models wrapped in constraints, tools, traces and evals that make failure visible before a caller finds it. The tools change; the loop stays.
telephony · ASR · barge-in
audio inLLM intent + deterministic guards
state machineAPIs · idempotency · auth
side effectsevals · traces · must / must-not
evidence outmodels
routing
knowledge
evals
agentic tooling
infra
// ai adoption
“One of our most AI-forward engineers.” — manager’s farewell
$ ./run-evals --suite silent-green currently: marketplace platforms at WayTrade · local inference on a 12 GB GPU · MCP servers · evals for AI-authored config
“Put deterministic logic where correctness matters. Put LLMs where judgment helps.”
— the rule I keep relearning
homelab
The homelab is where I test the ML I can’t put in production: local inference on a 12GB GPU, computer vision on nine cameras, and enough VLANs to make a network engineer smile. It is part curiosity, part skill-building, part just loving that it works.
RTX A2000 12GB passed through to the GPU VM. Ollama serves qwen3.5:9b, qwen3:8b-16k and nomic-embed-text; Whisper runs speech-to-text on CUDA; Open WebUI is the front door.
Frigate NVR with YOLOv9m on nine cameras, license-plate recognition on the driveway, and face recognition — all on the isolated camera VLAN.
Immich with CUDA-accelerated face and object recognition. Thumbnails on SSD, cache on a dedicated disk, originals on the NAS over NFS.
AMD Ryzen 5 2600X, 62GB RAM. Multiple VMs with VLAN isolation and PCIe passthrough. Dual network: 1Gbps plus a 2.5Gbps direct link to the NAS.
TerraMaster + OpenMediaVault. Separate volumes for media, personal files, VM backups and ISOs, exported over NFS with jumbo frames on the 2.5Gbps link.
Five VLANs, GL.iNet Flint 3, managed switches, Tailscale subnet routing and split DNS. Authentik SSO, Vaultwarden, AdGuard and CrowdSec keep the front door boring.
the lab is invite-only · authentik sso · access →
access
I run Jellyfin, Immich and a few other services for people I actually know. If you’d like access, reach out and I’ll set you up. One login, everything in one place.
by invitation
Request access
Jellyfin, Immich & more — accounts are set up personally.
secured via Authentik SSO — you’ll need credentials from me directly.