Measured writing on AI cost & risk
Every claim cited or measured. Field notes on governing agents, cutting spend, and the economics of AI-assisted engineering.
Introducing FORG v3: The AI Control Plane
Accurate per-model token accounting, co-primary adapters, and the signal layer for cross-tool intelligence — what shipped in v3 and why.
Read postThe FORG Index, Explained
How the FORG Index decomposes measured savings from estimated opportunity — the honest way to score AI efficiency.
Read postWhy Metadata-First Observability Is the Right Approach
Prompts and code never leave your machine. How metadata-only capture gives you cost and governance without the privacy tradeoff.
Read postBuilding Compliant AI Systems: SOC 2, GDPR & CCPA
A practical, honestly-caveated look at where AI coding tools intersect SOC 2, GDPR, and CCPA — and what FORG does and does not claim.
Read postFORG + Claude Code: Deep Integration Guide
Connect Claude Code to FORG for session telemetry, budgets, and per-model cost attribution — step by step.
Read postThe Hidden Cost of AI Coding Tools
Where AI spend actually goes on an engineering team, the common waste patterns, and the ROI of fixing them (illustrative).
Read postHow We Cut AI Costs 43% (Illustrative)
An illustrative, modeled walkthrough of the levers — downgrade routing, caching, and budget caps — that move an AI bill.
Read postRead the work, then start governing
Replace the estimates with real numbers from your own agents.
Start 14-day trial