Engineering deep-dives, product updates, and perspectives on observability, compliance, and cost optimization for teams shipping with AI.
A practical breakdown of how we identified waste, implemented budget rules, and measured the results. Includes before/after charts and the exact rule configs we used.
You don't need to store prompts to understand your AI usage. A deep dive into what metadata tells you, and why that's actually more valuable than payload logging.
Step-by-step guide to connecting FORG with Claude Code. Session tracking, cost attribution by developer, hooks configuration, and what signal data looks like.
The compliance landscape for AI tooling in 2025. What each framework requires, where they overlap, and how FORG's architecture helps you satisfy all three.
The average 20-person engineering team spends $3,200/month on AI tools they can't account for. Here's the breakdown and how to fix it.
FORG enforces k-anonymity ≥ 5 across all usage reports. What that means, how we implement it, and why it matters for your team's privacy.
Today we're launching two new plans designed for teams. Shared dashboards, org-level budgets, SAML SSO, and priority support.
A technical walkthrough of FORG's rules engine: rule types, evaluation order, conflict resolution, and budget enforcement with sub-millisecond overhead.
FORG Atlas embeds your usage patterns and lets you query them in plain English. How it works, what you can ask, and why we built it.
Breaking down the adapter protocol changes in v2: batching, compression, retry semantics, and the new session ID format.
Three fintech teams share how they use FORG to maintain audit trails, enforce data-handling policies, and satisfy their SOC 2 auditors.
FORG now supports SCIM provisioning and SAML 2.0 / OIDC SSO. How to configure it, what auto-provisioning looks like, and RBAC mapping.
Using 60 days of FORG usage data, we analyzed cost-per-task across Claude 3.5 Sonnet and GPT-4o across 8 task categories.
The FORG adapter protocol spec is now open source under MIT. Build your own adapter for any LLM tool, and we'll feature it in the community directory.