Sign in

Luke Murray

@luke.geek.nz
1.5K followers 910 following 1.3K posts

Microsoft Azure MVP 🌐 | MS Learn & Startup Advocate 📘 | Patterns & Practices & Cloud Native 🤖 | Coffee-Driven ☕️ My views are my own. He/Him. luke.geek.nz

PostsRepliesMedia
Luke Murray @luke.geek.nz · 22/09/2026
Delete Your Polling Loop: Change-driven programming in Node.js and Python with definitely NOT drasi/lib!
techcommunity.microsoft.com
Delete Your Polling Loop: Change-driven programming in Node.js with @drasi/lib | Microsoft Community Hub
In architectural discussions surrounding modern data systems, we previously examined how pivoting from an event-driven to a change-driven paradigm introduces...
010
Luke Murray @luke.geek.nz · 22/09/2026
Good day for a walk!
040
Luke Murray @luke.geek.nz · 17/09/2026
Join me tomorrow for Visual Studio Code watch party! #VSCode #visualstudio #mvpbuzz techcommunity.micros...
000
Luke Murray @luke.geek.nz · 13/09/2026
github.com/lukemurraynz...
github.com
GitHub - lukemurraynz/Groundwork: Groundwork: an AI-planned, deterministically-executed enterprise data platform provisioner for Azure
Groundwork: an AI-planned, deterministically-executed enterprise data platform provisioner for Azure - lukemurraynz/Groundwork
001
Reposted by Luke Murray
Dr Greg Low @greglow.bsky.social · 12/09/2026
Fabric RTI 101: Partitioning strategies blog.greglow.com/2026/09/12/f...
021
Luke Murray @luke.geek.nz · 06/09/2026
New toy has arrived - will see what I can do with it! #github #badge thegithubshop.com/collections/...
100
Luke Murray @luke.geek.nz · 05/09/2026
[Share] Foundry Labs The home for cutting-edge AI experiments from Microsoft
labs.ai.azure.com
Microsoft Foundry Labs | Early-Stage AI Experiments & Prototypes
Microsoft Foundry Labs is the home of cutting-edge AI experiments from Microsoft. Explore frontier-research models, interactive playgrounds and open-source releases — built today, available to you now.
000
Luke Murray @luke.geek.nz · 31/08/2026
Agent Package Manager AI coding tools are getting better quickly. The setup around them is still too often a pile of local files, copied prompts, one-off MCP servers, and tribal knowledge. #apm #mvpbuzz #github #agents #coding
luke.geek.nz
Agent Package Manager for AI Azure Engineering | luke.geek.nz
Agent Package Manager (APM) brings package.json-style manifests, lockfiles, and policy-based governance for AI agents in Azure engineering.
741
Luke Murray @luke.geek.nz · 31/08/2026
RAG that never goes stale: continuous vector sync with Drasi Every RAG pipeline I have looked at has embeddings generated at index time . I tested Drasi's SyncVectorStore reaction with Azure AI Search for real changes. #drasi #aisearch
luke.geek.nz
RAG that never goes stale: continuous vector sync with Drasi | luke.geek.nz
Keep RAG embeddings fresh with Drasi's continuous vector sync—per-change embeddings and deletions, lower cost than nightly reindexing.
010
Luke Murray @luke.geek.nz · 30/08/2026
Give your Copilot a live data feed with Drasi's MCP Reaction Ask an agent "what's the status of order 47?" and it answers from whatever it retrieved minutes ago. Drasi has an MCP Reaction that closes that gap.
luke.geek.nz
Give your Copilot a live data feed with Drasi's MCP Reaction | luke.geek.nz
Give Copilot a live data feed with Drasi's MCP Reaction: subscribe to continuous queries for instant push updates—no polling, real-time agent context.
020
Luke Murray @luke.geek.nz · 30/08/2026
Your cluster is a graph: drift detection with Drasi and Cypher Every AKS cluster running GitOps already has tools watching for drift. I tested Drasi queries over an AKS resource graph. #aks #drasi
luke.geek.nz
Your cluster is a graph: drift detection with Drasi and Cypher | luke.geek.nz
Detect cross-resource drift in AKS with Drasi and Cypher—standing Cypher queries over your cluster graph for temporal, relational compliance and alerting.
010
Luke Murray @luke.geek.nz · 30/08/2026
Detecting absence: querying for the events that never happened #mvpbuzz #drasi #aks
luke.geek.nz
Detecting absence: querying for the events that never happened | luke.geek.nz
Detecting absence with Drasi: continuous PostgreSQL queries (trueLater/trueFor) to alert when events stop—replace pollers with real-time, reliable alerts.
010
Luke Murray @luke.geek.nz · 28/08/2026
Downgrade a movie: Starship Poopers
000
Luke Murray @luke.geek.nz · 26/08/2026
Ohh GoT War for Westeros RTS! store.steampowered.com/app/2838950/...
store.steampowered.com
Game of Thrones: War for Westeros on Steam
Classic real-time strategy in the world of Westeros. Conquer the Seven Kingdoms solo, or dominate your enemies to claim the Iron Throne in multiplayer skirmish. Command the Great Houses, rally iconic ...
000
Luke Murray @luke.geek.nz · 25/08/2026
[Share] hve-agent-skills Just published some of my AI coding-agent skills for Azure platform work. Each skill encodes real deployment experience: traps, verification gates, and decision rather than tutorials. #aks #drasi #sreagent #rust
github.com
GitHub - lukemurraynz/hve-agent-skills: Production-hardened AI coding-agent skills for Azure platform work
Production-hardened AI coding-agent skills for Azure platform work - lukemurraynz/hve-agent-skills
010
Luke Murray @luke.geek.nz · 25/08/2026
0x Alpha Free, not too bad. Some say its GLM 5.3 Vision - but it seems to be doing alot of the work I need to do.
100
Luke Murray @luke.geek.nz · 24/08/2026
📢 Microsoft Global Agent‑a‑thon 🌐 Online | Multiple session times 📅 Sep 17, 2026 🔗 Register here: aka.ms/AgentathonSep...
100
Luke Murray @luke.geek.nz · 22/08/2026
When the AI overlords take over the digital net, I can sit in my bunker with my SEGA.
2203
Luke Murray @luke.geek.nz · 20/08/2026
Evidence to Decision: building an AI-envisioning workshop tool on Azure Most "AI workshop" tooling is either a whiteboard with sticky notes or a slide deck. I built something different. #mvpbuzz #azure #aienvisioning
luke.geek.nz
Evidence to Decision: building an AI-envisioning workshop tool on Azure | luke.geek.nz
How I built a governed, evidence-first AI workshop pipeline on Azure, from sticky notes to approved delivery documents, with every step auditable.
010
Luke Murray @luke.geek.nz · 19/08/2026
KubeFleet Community call ANZ is on this evening!! KubeFleet is a cloud-native solution for multi-cluster application management! #kubefleet #mvpbuzz #opensource #aks 5.30 - 6.30 PM NZT!
github.com
GitHub - kubefleet-dev/community: Governance and community material for KubeFleet · GitHub
Governance and community material for KubeFleet. Contribute to kubefleet-dev/community development by creating an account on GitHub.
020
Luke Murray @luke.geek.nz · 15/08/2026
Azure Feeds Weekly: 08 Aug 2026 to 15 Aug 2026 is out! azurefeeds.bmeurl.co...
000
Luke Murray @luke.geek.nz · 14/08/2026
🔔 Azure Infrastructure Resiliency Manager + Resiliency Agent 🟢 Start resilient: guidance + Bicep 🔵 Get resilient: assess + remediate 🟣 Stay resilient: detect drift + prove recovery IRM is in public preview; Agent rollout completes Aug 18.
learn.microsoft.com
Infrastructure Resiliency Manager overview - Azure
Learn about Infrastructure Resiliency Manager in Resiliency in Azure and how to view resiliency posture across your resources and service groups at scale.
000
Luke Murray @luke.geek.nz · 13/08/2026
Azure Monitor SLIs and the Azure SRE Agent I wanted to see whether an Azure Monitor SLI could give the Azure SRE Agent enough context to investigate a AKS failure. So I built a small service, measured it through Service Group scoped SLIs, and broke it.
luke.geek.nz
Azure Monitor SLIs and the Azure SRE Agent | luke.geek.nz
A practical walkthrough of Azure Monitor SLIs on Service Groups, error budget burn alerts, and Azure SRE Agent remediation for an AKS workload.
000
Luke Murray @luke.geek.nz · 11/08/2026
AKS Desktop is cool! #aks #headlamp learn.microsoft.com/...
010
Luke Murray @luke.geek.nz · 11/08/2026
My epic era of PC gaming! Still some of my faves.
120
Luke Murray @luke.geek.nz · 11/08/2026
I completely forgot I had this! Found while cleaning the cupboards. Microsoft Ignite bag, from NZ 2016! My one and only ignite, fond memories of it, and never thought years later I would be part of the MVP ecosystem! #mvpbuzz #msignitenz
010
Luke Murray @luke.geek.nz · 04/08/2026
When you use Markdown so much you start typing Markdown into Word documents automatically, without thinking! _(AUTOPILOT!)_
121
Luke Murray @luke.geek.nz · 04/08/2026
Strangler fig: retiring a legacy dispatch layer #patterns #practices #stranglerfig
luke.geek.nz
Strangler fig: retiring a legacy dispatch layer | luke.geek.nz
Apply the strangler fig pattern to retire a legacy dispatch layer: domain-by-domain migration, route parity tests, ported tests, and safe cutovers.
000
Luke Murray @luke.geek.nz · 04/08/2026
Building Execution Ledger - Durable Workflow Orchestration in Rust on Azure Container Apps Execution Ledger is an open-source durable workflow orchestrator I built in Rust on Azure Container Apps. #rust #aca #mvpbuzz #drasi #azure
luke.geek.nz
Building Execution Ledger - Durable Workflow Orchestration in Rust on Azure Container Apps | luke.geek.nz
Learn how I built Execution Ledger, a durable workflow orchestrator in Rust running on Azure Container Apps with step-level replay, idempotency, compensation, KEDA autoscaling, and managed identity.
010
Luke Murray @luke.geek.nz · 03/08/2026
Execution Ledger built on Rust with Drasi support. Durable job orchestration for Azure. Step-level idempotency, replay, skip, and compensate controls. Tenant-scoped API with Entra ID auth. Connector packs for REST, queue, blob, SFTP etc.
github.com
GitHub - lukemurraynz/Execution-Ledger: Execution Ledger
Execution Ledger. Contribute to lukemurraynz/Execution-Ledger development by creating an account on GitHub.
020
Luke Murray @luke.geek.nz · 30/07/2026
When Your AI Agent Lies: Silent LLM Fallbacks A system I was debugging had AI-generated content flowing through it for days. The responses came back fast. No errors in the logs. No alerts fired. Every single response was templated.
luke.geek.nz
When Your AI Agent Lies: Silent LLM Fallbacks | luke.geek.nz
Detect silent LLM fallbacks: when AI agents return templated responses with 200 OK, monitor token usage and fallback rates to avoid hidden correctness failures.
510
Luke Murray @luke.geek.nz · 30/07/2026
Treat prompt changes like code deploys Most teams running LLM-backed features would never ship a code change without a test suite and a CI gate. Then they tweak a prompt, eyeball one output, and send it straight to production.
luke.geek.nz
Treat prompt changes like code deploys | luke.geek.nz
Prevent silent regressions by treating prompt edits like code deploys, with automated tests, CI gating, and continuous evaluation.
120
Luke Murray @luke.geek.nz · 30/07/2026
Avoiding Reasoning Model Failures with Microsoft Foundry Swapping only the deployment name is not a safe change for reasoning-tier models. Even with the same endpoint and SDK, reasoning deployments can fail in at least four different ways.
luke.geek.nz
Avoiding Reasoning Model Failures with Microsoft Foundry | luke.geek.nz
Learn four ways reasoning-tier models can break Azure OpenAI integrations and how to fix API versioning, params, token budgets, and reasoning_effort.
010
Luke Murray @luke.geek.nz · 29/07/2026
🥇 AzureRM Provider v5.0.0 dropped! * RP registration defaults changed (none not legacy) * Enhanced validation moved into features block * Preflight validation at plan time (opt-in) * Removed resources, deprecated flags #terraform #azurerm
github.com
Releases · hashicorp/terraform-provider-azurerm
Terraform provider for Azure Resource Manager. Contribute to hashicorp/terraform-provider-azurerm development by creating an account on GitHub.
000
Luke Murray @luke.geek.nz · 25/07/2026
This is neat! Dragons! www.glasgowlive.co.uk/news/glasgow...
glasgowlive.co.uk
Game of Thrones dragon soars over Glasgow to celebrate Commonwealth Games opening
The eight metre flying model took to the skies again this morning.
120
Luke Murray @luke.geek.nz · 25/07/2026
My first Tauri app! Its not perfect but its bloody fast. v2.tauri.app
000
Luke Murray @luke.geek.nz · 25/07/2026
Saturday morning means brunch.
010
Luke Murray @luke.geek.nz · 24/07/2026
Microsoft Foundry hosted agents: lessons learned building a production multi-agent service Some of my thoughts and findings on my first deployment of Foundry Hosted Agents.
luke.geek.nz
Microsoft Foundry hosted agents: lessons learned building a production multi-agent service | luke.geek.nz
Lessons and best practices for hosting production multi-agent services with Microsoft Foundry hosted agents—compute, protocols, sizing, and deployment tips.
130
Luke Murray @luke.geek.nz · 23/07/2026
Nastiest bug of the month: AppConfigurationClient configured with trim_prefixes while every reader queried fully-namespaced keys. No errors - every single config read just silently returned the default. Config mismatches like this fail quiet. Log your resolved values.
010
Luke Murray @luke.geek.nz · 23/07/2026
I got a weird look when I told someone last week, I was refactoring something to RUST! Now I need to actually learn it!! Time to sleeves up, thanks @willvelida.com! rustup time!
111
Luke Murray @luke.geek.nz · 22/07/2026
Reasoning-tier gpt-5 models (gpt-5, 5.1, 5.5, o1, o3-mini) reject temperature and most sampling params outright. gpt-5-chat still accepts it though. Make sampling params optional with drop-and-retry instead of hardcoding them.
000
Luke Murray @luke.geek.nz · 22/07/2026
Gotcha: Azure OpenAI's /openai/v1 Responses API doesn't take literal "v1" or "preview" as api-version — it pins to a moving preview date-string (now 2025-11-15-preview). Older date-style versions 400 immediately. Verify the current value before hardcoding it.
010
Luke Murray @luke.geek.nz · 22/07/2026
Assistants API hard-retires 2026-08-26 on OpenAI AND Azure OpenAI - same deadline. But migration paths differ: OpenAI devs move to the Responses API; Azure OpenAI devs move to Foundry Agent Service (built on Responses). Follow Microsoft's guide, not OpenAI's.
000
Luke Murray @luke.geek.nz · 22/07/2026
Azure SDK clients refuse bearer tokens over plain HTTP. If you're capture-replaying Azure SDK calls against a local mock server, your test doubles need AzureKeyCredential, not DefaultAzureCredential - otherwise it's not the mock failing, it's the transport.
000
Luke Murray @luke.geek.nz · 21/07/2026
Linking a Container App as a Static Web Apps backend caps you at a hard 45s timeout and HTTP-only - no WebSocket, no matter what your Container App supports. Long-running or persistent-connection agent endpoints? Go through Front Door or hit the Container App directly.
010
Luke Murray @luke.geek.nz · 21/07/2026
ACA gotcha before promising compliance a checkbox: there's no CMK setting for the managed environment itself. You can CMK dependent services (ACR images, mounted Files) but not the environment. Mandated CMK on the compute platform? That's an AKS conversation.
110
Luke Murray @luke.geek.nz · 21/07/2026
Building an MCP server on top of an internal agent framework? Don't hand-pick a subset of tools for MCP exposure - route the whole tool registry through the same audit/authorize decorators your other surfaces use, so MCP clients see the same capability set as your other clients.
020
Luke Murray @luke.geek.nz · 21/07/2026
TIL (the hard way): Container Apps ingress timeout defaults to 240s but isn't fixed — you can raise it up to 2400s (40 min) at the environment level via --ingress-timeout.
110
Luke Murray @luke.geek.nz · 21/07/2026
If prompts are part of your deployable surface, treat "promote to prod" like a merge to main: gate it on eval pass, not vibes. We added a hard PermissionError if the latest eval run for a prompt failed or never ran.
000
Luke Murray @luke.geek.nz · 20/07/2026
Debugging tip: if your agent pipeline has a deterministic fallback path, watch for it firing silently. Our tell: response time <5s + total_tokens_used: 0 = the LLM call never actually happened. Three stacked misconfigs caused this & every symptom looked like a working agent.
010