Alpha Dashboard

Overview

The Alpha Dashboard is a browser-based monitoring interface that visualizes real-time Hyperliquid market microstructure. It connects to the Unified Stream and renders six signal panels, each displaying a different dimension of on-chain and off-chain activity. Together, these panels provide a composite view of market state that spans execution flow, network health, and cross-venue dynamics.

The dashboard is available to all Pro tier subscribers at /alpha/live.

Panel Architecture

The dashboard is organized into six panels. Each panel consumes one or more event types from the Unified Stream and renders a specialized visualization. Panels are cross-linked — events in one panel can trigger highlights in others, enabling rapid correlation of multi-dimensional signals.

1. Liquidation Radar

Displays ABI-decoded HyperCore RawAction liquidation events in real time. Each event is parsed from the on-chain limit-order payload to extract the actual side (buy/sell), liquidation price, and position size.

The server-side cascade detection model operates directly on the hypernode, applying a rolling-window threshold to identify when multiple liquidation events cluster within a configurable time interval. When the event count within the window exceeds the threshold, a liquidation_cascade signal is emitted.

Key metrics displayed:

• Individual liquidation events with side, price, and size
• Cascade detection status and active cascade count
• Cumulative liquidation volume over rolling windows (1m, 5m, 15m)
• Heatmap of liquidation density by price level

2. Order Flow Pressure

Visualizes directional execution pressure using a decomposition of trade flow into buy-initiated and sell-initiated volume. The panel provides:

• Buy/sell split bar — Proportional breakdown of aggressive flow by side.
• Net delta tracker — Cumulative signed volume $\Delta(t) = \sum_{i} q_i \cdot \text{sign}_i$ over configurable lookback windows, where $q_i$ is trade size and $\text{sign}_i \in \{+1, -1\}$ indicates buyer- or seller-initiated.
• Orderbook imbalance — Ratio of bid-side to ask-side depth within a configurable number of levels from the best quote. An imbalance ratio significantly different from 1.0 indicates directional pressure building in the limit order book.

This panel helps identify whether aggressive flow is concentrated on one side of the book, signaling potential short-term directional moves.

3. Block Activity Monitor

Tracks HyperEVM block production metrics including block size, gas utilization, and inter-block cadence. The panel includes a consensus pulse regime analysis that compares base fee dynamics between different block timing regimes:

• Fast blocks (inter-block time $\leq 3$s) — typically low-gas utility blocks during high-throughput periods.
• Slow blocks (inter-block time $\geq 20$s) — potentially indicating validator latency or reduced network activity.

The base fee delta between these regimes (measured in gwei) serves as a proxy for network load distribution.

Key metrics displayed:

• Block height and gap since last block
• Gas utilization percentage per block
• Regime base fee deltas (fast vs. slow blocks, in gwei)
• Tokyo P2P latency as a measure of network propagation health

4. Market Stress Gauge

A composite regime signal that synthesizes inputs from all other panels into a single scalar stress score. The gauge is computed as:

$S(t) = \sum_{i} w_i \cdot f_i(t)$

where $f_i(t)$ are normalized feature inputs derived from each panel’s data stream, and $w_i$ are adaptive weights that adjust based on prevailing market regime. Each $f_i$ is mapped to the unit interval $[0, 1]$ via a normalization function calibrated to historical quantiles, ensuring that the composite score $S(t)$ remains interpretable across different volatility environments.

The specific features $\{f_i\}$, weight adaptation logic, and normalization calibration parameters are proprietary. The dashboard displays:

• The composite stress score $S(t)$ on a gauge dial (0 = calm, 1 = extreme stress)
• A contributing factor breakdown showing which panels are driving the current reading
• Historical stress trajectory over a configurable lookback window

The stress gauge is designed as a situational awareness tool — it highlights periods where multiple microstructure dimensions are simultaneously signaling unusual activity.

5. Signal Timeline

A chronological correlation rail that cross-highlights events across all source panels. The timeline renders a unified event stream with synchronized timestamps, enabling visual correlation of causally related events.

When a liquidation cascade fires, the corresponding block activity changes (gas spikes, cadence shifts) and order flow regime transitions (delta reversals, imbalance shifts) are highlighted in the same time window. This cross-panel linking is automatic and driven by temporal proximity.

Interaction model:

• Click any event on the timeline to highlight correlated events across all panels.
• Drag to select a time range and filter all panels to that window.
• Hover over a signal to see its full metadata in a tooltip.

6. Order Book Depth

Full L2 bid/ask ladder showing per-level size, cumulative depth, and live spread. The visualization updates in real time from the Unified Stream’s depth channel.

Key metrics displayed:

• Per-level bid and ask quantities
• Cumulative depth curves (bid and ask)
• Spread in basis points and absolute terms
• Depth imbalance ratio at configurable levels (top 5, top 10, top 20)

Data Source Architecture

The dashboard consumes data from two distinct layers within the Aleatoric pipeline.

Layer 1 — Raw Transport (L1)

The raw transport layer provides direct access to protocol-level data from Hyperliquid and reference venues:

Transport	Use Case
JSON-RPC	On-demand queries for snapshots, metadata, and historical lookups
WebSocket	Continuous price feed for L2 book updates and trade ticks
gRPC	Lowest-latency streaming for latency-sensitive panel updates
Hyperliquid /info	Instrument metadata, open interest, and funding rate snapshots

Layer 2 — Pre-Decoded Intelligence (L2)

The intelligence layer processes raw data into structured, normalized event streams consumed by the dashboard panels:

Feed	Description
Unified Event Stream (SSE)	Primary data bus for all panel data; multiplexes all event types on a single connection
Liquidation feed	ABI-decoded liquidation events with cascade annotations
Order flow feed	Trade flow decomposition with directional classification
Block activity feed	Block metrics, gas analysis, and consensus pulse data
Stress feed	Composite stress gauge values and contributing factor breakdowns
Market snapshot helpers	Point-in-time state snapshots for panel initialization on connect

All L2 feeds use normalized schemas. Event types share common fields (timestamp, coin, event_id) and type-specific payloads, enabling uniform parsing across the dashboard frontend.

Access Tier Matrix

The Alpha Dashboard is included with every Pro subscription. No additional configuration is required — authenticate with your Aleatoric account and the dashboard connects automatically.

Tier	Dashboard Access	Data Retention	Export
Free	—	—	—
Basic	—	—	—
Pro	Full dashboard access	24-hour lookback	CSV export of visible window
Quant	Full dashboard access	90-day lookback	CSV + API export, webhook alerts

Performance Considerations

• The dashboard opens a single SSE connection to the Unified Stream, multiplexing all event types. This counts as one concurrent stream against your tier’s rate limit.
• Panel rendering is throttled to 60 fps to prevent browser overload during high-activity periods (e.g., liquidation cascades with hundreds of events per second).
• Historical lookback queries (for chart initialization and time range selection) use the JSON-RPC endpoint and count against your standard rate limit.

Next Steps

• Unified Stream — The underlying event feed powering the dashboard
• Divergence Radar — Cross-venue arbitrage signal detection
• Python SDK — Consume the same data programmatically
• TypeScript SDK — Build custom frontends with typed events
• Connecting — Integration examples for each transport protocol
• Rate Limits — Understand throughput constraints