> For the complete documentation index, see [llms.txt](https://docs.yala.org/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.yala.org/prediction-market-intelligence/late-stage-multi-agent-fair-value-system.md).

# Late Stage: Multi-Agent Fair-Value System

In the late stage, Yala evolves into a comprehensive multi-agent fair-value system.

### Expanded Inputs

* Any asset or event category
* Broader prediction-market landscape (crypto, equities, elections, esports, macro)
* Future time horizons

### Outputs

* Probability density functions (PDFs)
* Multi-factor fair-value curves
* Confidence intervals
* Distribution shapes

### Supervisor and Worker Agent Architecture

**Supervisor (Orchestrator) Agent**

* Parses user queries
* Assigns tasks to Worker Agents
* Monitors execution
* Aggregates outputs into final results

**Worker Agents**

* Fair-Value Modeling Agent
* Data Collection Agent
* Sentiment Analysis Agent
* Smart-Money Analysis Agent
* Event Tracking Agent
* Options Analysis Agent
* Simulation Agent
* Decision Aggregation Agent

Together, these agents generate explainable, multi-factor fair-value assessments.

### Private-Information Adjustment

The late stage introduces:

* Insider / Private-Information Adjustment Agent
* Encrypted vector storage
* Confidential RAG pipelines

This allows users to securely contribute private signals that refine subjective fair values.

<br>