技術的ディープダイブ
階層理論ナビゲーション、フェーズコンテキスト切替、モデルオーケストレーション戦略、評価メトリクス — エンジニアと研究者のために。
Deep Dive
A Layered Theory of Decision Engine OS
Dive as deep as your role requires.
Built to let organizations survive AI.
Adapt to Context. Never Compromise Principles.
Organizations change priorities over time. MARIA OS adapts automatically.
Balanced optimization active.
Change priorities without changing who you are.
Model Architecture
Fast × Heavy Model Orchestration
Models are separated by role, not capability.
That's why we separate models by role, not by power.
Fast Model
PreprocessingHeavy Model
ReasoningDecision Flow
ScrollWhy separation enables stopping
Smart models never execute
Fast models never judge
Authority lives outside models
Human gates are enforced
Neither model is permitted to execute.
Control architecture, not model architecture.
How to Measure a Decision OS
Can it reliably stop decisions that should not be executed?
Why MARIA OS Can Stop
Spec Sheet
An OS that controls whether decisions should be executed.
Control Logic & Implementation Design
Not ML evaluation — control systems engineering.
5 Independent Evaluation Layers
completeness < threshold → STOPVerify that all premises required for judgment are in place and not contradictory.
State Machine Control
1// State Machine - Decision Control2enum DecisionState {3 AUTO_EXECUTE, // Full automation4 HUMAN_REVIEW, // Requires human check5 CEO_REVIEW, // Escalate to CEO6 BOARD_REVIEW, // Board-level decision7 STOP // Halt execution8}910// Models have NO execution authority11// Only OS layer manages state transitions12function evaluateDecision(context: Context): DecisionState {13 const premise = checkPremiseConsistency(context);14 const stability = checkDecisionStability(context);15 const impact = checkImpactIrreversibility(context);16 const philosophy = checkPhilosophyDeviation(context);17 const explain = checkExplainability(context);18Fast / Heavy Separation
Premise org, state updates, deterministic
Deep reasoning, on-demand, high-cost
Neither model has execution authority.
A control OS that prevents decisions from breaking.
From Judgment to Scalable Autonomy
AI is already autonomous.
The question is: can your organization trust it?
What organizations face today:
Not an AI problem.
The absence of structured judgment.
Not about making AI smarter.
It is not "a platform to optimize AI agents."
Condense judgment into structures
Transplant into an operating system
Scale safely through autonomous agents
That is the essence of MARIA OS.
Autonomy scales.
Responsibility does not.
Up to 10,000 AI agents per organization. But execution only scales.
Decision points
→ Decision Axis
Approvals
→ Responsibility Gates
Human decisions
→ "human-sized"
Execution scales.
Responsibility stays human-sized.
AI operations run without exhaustion, without breaking, with full replicability.
Universe Builder creates
replicable AI orgs.
Not just a group of agents.
Universe =
Complete operational unit: judgment, responsibility, optimization, approval.
Universe Builder flow:
Once working, replicate across departments and environments.
HITL is not a fallback.
A collaboration point.
Not a mechanism to compensate for AI failure.
Designed from the start:
Meaningful decisions
Quality assurance
Learning assets
AI and humans collaborate, not compete.
Why traditional approaches fail
| Aspect | Traditional | MARIA OS |
|---|---|---|
| Judgment | Implicit | Explicit Decision |
| Responsibility | Person-dep. | Gates |
| Optimization | Everything | Surface only |
| Values | Documents | Constraints |
| Reuse | Difficult | Per-Universe |
| Audit | After fact | Always on |
| Scale | Chaotic | Stable |
How organizations transform
Before
After
Not an improvement.
A transformation.
What is MARIA OS?
Elevates judgment to OS
Locks values as constraints
Scales autonomy safely
Human-AI collaboration
MARIA OS is a
Decision Engine OS
Does not automate judgment.
Makes it scalable, governable, reusable.