AI Allocation Engine

Purpose

The allocation engine is responsible for deciding where and how staked SOL is deployed to achieve optimal risk-adjusted returns.

The system is designed to:

• Maximize sustainable yield

• Limit downside risk

• Adapt to changing network and incentive conditions

Data Inputs

The system evaluates multiple data categories, including:

• Base staking APY

• MEV rewards and distribution models

• Incentive programs (tokens, points, or additional rewards)

• Validator uptime, commission, and slashing history

• Liquidity and withdrawal constraints

• Network congestion and volatility

All data inputs are normalized to enable fair comparison across strategies.

Risk-Adjusted Scoring Model

Why Risk Adjustment Matters

Headline APY alone does not reflect real performance or safety.

LSD prioritizes net, risk-adjusted returns rather than short-term yield spikes.

Scoring Components

Each staking opportunity is evaluated using:

• Expected return

• Historical performance stability

• Smart contract and protocol risk

• Validator reliability

• Liquidity and exit latency

These factors produce a composite score used for allocation decisions.

Continuous Feedback & Learning

Performance Evaluation

After each allocation cycle:

• Actual realized returns are compared against projections

• Deviations are analyzed

• Underperforming strategies are deprioritized