Engineering
Notes
Writing from the Nodeblue team on software systems, industrial automation, AI deployment, and applied research. Direct, technical, no filler.
12
Articles
5
Topics
13m
Avg read
Covering
How we integrate AI into businesses — and what that actually looks like in practice.
All Posts
What Your EHR Doesn't Cover: Building Software Around Clinical Operations
EHR systems handle clinical documentation. They were never designed to run the operational layer around it. Here's where the gaps are and how custom software fills them.
Jessica Maria
April 24, 2026
What Agentic AI Actually Looks Like on the Plant Floor
AI agents operating inside industrial systems aren't the same as the chat assistants and document tools most teams are familiar with. The constraints are different, the failure modes matter more, and the architecture has to reflect that.
Dylan McCarthy
April 15, 2026
The US Manufacturing Gap Is a Tooling Problem
American engineers are not less skilled than their counterparts in Chinese factories. The facilities they work in are less capable. That difference matters because it points to a different solution.
Dylan McCarthy
April 13, 2026
Building Production AI Agents: What the Demo Doesn't Show
The gap between a working prototype and a reliable agent running in production is wider than most teams expect. Here's what breaks, and how to design around it.
Brandon Sheedy
April 12, 2026
Offline-First Apps for Field and Warehouse Operations
Most business software breaks the moment connectivity drops. For teams working in warehouses, field sites, or anywhere with spotty cellular, that's not an edge case — it's the job. Here's how to build software that handles it.
Stephen Pusateri
April 8, 2026
OPC-UA vs MQTT: Picking the Right Industrial IoT Protocol
Both protocols appear on every IIoT architecture diagram. Picking the wrong one costs you in ways that don't show up until you're in commissioning. Here's how to decide.
Dylan McCarthy
April 5, 2026
How We Integrate AI Into Existing Business Software
Most businesses don't need to rebuild their software to add AI. They need AI that understands their specific data, respects their access controls, and connects to the tools their teams already use.
Dylan McCarthy
April 1, 2026
The Quiet Cost of Inherited PLC Code
Undocumented PLC programs don't announce their debt. It accumulates silently until a process change, a personnel transition, or a fault exposes exactly how much institutional knowledge is locked inside a ladder diagram.
Brandon Sheedy
March 28, 2026
Why Workflow Automation Breaks Down in Year Two
The automation that saved your team 20 hours a week in the first year becomes a maintenance burden by the second. The pattern is consistent and avoidable.
Stephen Pusateri
March 18, 2026
LLM Evaluation Beyond the Benchmark
Benchmark scores predict benchmark performance. They are a weak signal for production quality on your specific task. Here's how to build an evaluation harness that actually tells you something useful.
Dylan McCarthy
March 8, 2026
Securing Industrial Control Systems: What Actually Works
OT security frameworks are easier to audit than to implement in a plant that hasn't shut down in six years. Here's what separates the guidance that holds up from what gets deferred indefinitely.
Brandon Sheedy
February 24, 2026
Have a project in mind?
We build software, automation systems, and AI tools for businesses that need them to actually work.