Skip to content
View rduffyuk's full-sized avatar
0 followers · 12 following
  • UK
  • 11:20 (UTC +01:00)

Achievements

Achievement: Pair Extraordinaire

Achievements

Achievement: Pair Extraordinaire

Block or report rduffyuk

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
rduffyuk/README.md

Ryan Duffy

SRE specialist building Rootweaver — a personal AI infrastructure platform on K3s, with prosumer GPU, autonomous SRE triage agent, RAG-backed knowledge graph, and MCP-as-primary surface. Solo build, ~12 months of evenings + weekends.

Day job: SRE at BT Group (UK, 14-year tenure) Side build: Rootweaver — single-engineer end-to-end personal AI platform

Read more

What I work on

  • K3s + vLLM + Qdrant + FalkorDB — production homelab AI stack
  • MCP aggregator pattern — 29 tools spanning vector search, code-intelligence graph, workflow analytics
  • Autonomous SRE agent — Alertmanager → Kafka → LLM-backed triage with 5-way evidence fan-out
  • GitOps with FluxCD, observability with Prometheus + Grafana + Tempo, sealed-secrets
  • AI orchestration as a primary engineering mode — directing Claude Code / Aider / Perplexity as a virtual team

Pinned series

Latest: Season 3: Platform Engineering in Public — Rebranding from ConvoCanvas to Rootweaver, autonomous SRE agent, streaming journals with Kafka + vLLM, dual knowledge graphs.

Popular repositories Loading

  1. querymind querymind Public

    🧠 Intelligent RAG with smart query routing - Choose the right search strategy automatically (FastSearch <1s, DeepResearch ~10s, WebSearch 2-5s)

    Python 1

  2. convocanvas-landing convocanvas-landing Public

    ConvoCanvas coming soon landing page

    HTML

  3. blog-astro blog-astro Public

    Ryan Duffy's blog - AI Infrastructure & Local LLM Journey (Astro 5.0 + Cloudflare Pages)

    Astro

  4. sig-release sig-release Public

    Forked from kubernetes/sig-release

    Repo for SIG release

    Shell

  5. rduffyuk rduffyuk Public

    Ryan Duffy — SRE building Rootweaver

  6. engineering-memory-benchmark engineering-memory-benchmark Public

    Empirical study: layered retrieval (typed→semantic→grep) scores 0.954 for LLM-generated engineering artifacts. 5 conditions, 3 model tiers, 36 generated ADRs, 23 score files.

    Python