Index/ AESOP/ AESOP Transformation OS
AESOP Studio

AESOP Transformation OS

Discover → Design → Build → Evaluate → Repair → Track — organization-scale agent development with portfolio management, org context ingestion, and strategic prioritization

Discover Design KB Audit Build Evaluate Repair
01
Stage One
Discover
Structured problem-first interview. Validate the opportunity before writing a single instruction.
  • Guided intake interview
  • Problem validation
  • Opportunity brief
  • Platform fitness check
02
Stage Two
Design
Generate a full PRD and get architecture recommendations across platforms.
  • PRD generation
  • Architecture scoring
  • Platform selection
  • Feasibility review
03
Stage Three
KB Audit
Scan your knowledge base for quality issues before they corrupt agent responses.
  • Per-doc quality scoring
  • Gap analysis
  • Conflict detection
  • Markdown report export
04
Stage Four
Build
Phase-by-phase instruction writing with human-in-the-loop checkpoints and artifact export.
  • 5-phase build pipeline
  • HITL checkpoint
  • Platform artifacts
  • JSON schema export
05
Stage Five
Evaluate
Multi-agent scoring, red-team testing, bias probes, and certification across 6 dimensions.
  • 20+ evaluation agents
  • Red-team (Cave)
  • Bias probe
  • Platinum certification
06
Stage Six
Repair
AI-powered instruction rewrite using full evaluation context. Tracks regression across versions.
  • Extended thinking rewrite
  • Targeted patches
  • Regression detection
  • Version history
The AESOP Methodology — Problem First, Always
Problem PrimacyNo tool is selected before the problem is fully understood.
Evidence Over IntuitionEvery architectural decision is tied to validated user signals.
Build to MeasureInstructions without an evaluation plan are hypotheses, not agents.
Versioned TruthEvery change is tracked. Every regression is detectable.
Platform AgnosticThe best platform is the one that fits the problem, not the team's default.
Repair as RitualEvaluation is only valuable if it drives improvement.