Engineering Insights
Articles and nuggets on systems engineering, AI governance, and the architecture of complexity.
AI code isn't the problem. How you run it is.
The problem usually isn't the model. Directing a large, fast team of AI agents is a management and process skill, and review, tooling and the quality loop have to change with it.
Read Article
Good Managers Get Leverage. Bad Managers Meet the Gravity Well.
Agents run your brief at speed and often won’t flag a bad one. A clear brief is where the leverage starts, and checks the system can run are what make it hold.
Read Article
Defect-Free by Design: What Michael Fagan Taught Me About AI Agents
Every defect is a symptom, the disease is upstream. Learn how we apply the Fagan Inspection Process to create self-improving agentic production systems.
Read Article
Graph Engineering: WBS, DAGs, and Agentic Teams
Prompt, context, harness, and graph engineering are concurrent requirements for solid workplans, for humans and agents. The WBS and the task DAG are how that structure shows up.
Read Article
A Front Door for Founder Submissions, Built with SAVA-P
SAVA-P needed a front door for founder submissions, and Pairti built SAVA-F with them. These screens come from Pairti’s own workspace, running a submission from Knox, one of Pairti’s own portfolio companies.
Good Managers Get Leverage. Bad Managers Meet the Gravity Well.
Agents run your brief at speed and often won’t flag a bad one. A clear brief is where the leverage starts, and checks the system can run are what make it hold.
SIGO (S#*t In, Good Out): Structured PDF Analysis for AI-Native Applications
If your software has to act on badly laid out regulatory PDFs, you need structure you can trust before any model reasons over them. See what held up when we compared plain text, Docling, cloud vision models and a small model on a laptop, and where each one still slips.
The Harness Was Isolation. Benchmarks Are Eating It.
Isolation and obedience used to live in the harness you could audit. Scoreboards that reward self-contained agents are pulling both into opaque model state. Keep rails you can still see.
Graph Engineering: Why Your WBS Is the Right Model for Agentic Teams
Prompt, context, harness, and graph engineering are concurrent requirements for solid workplans, for humans and agents. The WBS and the task DAG are how that structure shows up.
Defect-Free by Design: What Michael Fagan Taught Me About AI Agents
Every defect is a symptom, the disease is upstream. Learn how we apply the Fagan Inspection Process to create self-improving agentic production systems.
JOMC: Why AI Agents Can’t Refactor Themselves
AI agents are world-class pattern matchers, but they lack architectural wisdom. Learn why humans must act as the "Circuit Breaker" to prevent the recursive loop of fragile fixing.
The Physics of AI Orchestration
Orchestrating agents looks a lot like orchestrating massive programs of work. How to use Systems Engineering principles to master AI.
The Foreman’s Desk: Schedule Rigor in the Age of AI
Agile task lists fail at AI scale. Learn how we use BP6 to bring program scheduling discipline and Critical Path analysis to the "AI Excavator."
The Padded Room: Engineering a Safe Space for AI Agents
AI models are confident but prone to hallucination. Learn how we use strict internal standards and "padded room" configurations to turn chaotic potential into reliable engineering.
The Re-Awakening of Systems Engineering
We are moving from "spec-then-code" to an iterative feedback loop where AI helps refine requirements. Why disciplined specification is the new superpower in the age of AI.
Nuggets
Short, sharp technical notes, same Blog, tighter form.
