What It Means
Taxonomy config is how you turn 'our brand standards' and 'our compliance rules' into machine-enforceable rubrics — without writing code.
Taxonomy configuration is the layer that turns your requirements into enforceable rules for the AI QA & Evaluation Platform. Instead of building custom code for each compliance domain or brand standard, you express your rules as configuration: label definitions (what safe_to_deploy / needs_fix / blocked mean for your context), scoring criteria (what triggers each verdict), failure categories (types of issues your rubric catches), escalation thresholds (when to route to SME through authority escalation), and required evidence fields. The taxonomy is version-controlled through rubric versioning, so every review decision can be traced to the specific rules that applied at the time. New project? Reconfigure the taxonomy. Different compliance domain? New taxonomy. No custom engineering required — which means pilots launch in days, not weeks.
Why It Matters
Every organization has different rules. A financial services firm following FINRA 2210 has completely different requirements than a B2B SaaS company worried about product accuracy. Without configurable taxonomies, you'd need to rebuild the review system for each customer. With them, you reconfigure. And because taxonomies are version-controlled, the immutable audit trail always shows exactly which rules applied to each historical decision — even as rules evolve.
How Bookbag Helps
Bookbag's taxonomy system lets you define rubrics, label sets, scoring criteria, and escalation rules as configuration — no code required. The annotation interface dynamically adapts to the active taxonomy, showing reviewers the right fields, criteria, and escalation options for their specific project. Every review decision references the taxonomy version used. Template taxonomies for common use cases (FINRA compliance, B2B SaaS outbound, cold email quality) give you a starting point you can customize.
Related Terms
Frequently Asked Questions
Related Resources
Solutions
Compare
See comparison →See how Bookbag works
Join the teams shipping safer AI with real-time evaluation, audit trails, and continuous improvement.