phase 14-15: docker build validation and helm deployment

2026-04-11 11:59:45 -07:00
parent 7394d241c9
commit ce10afa034
179 changed files with 32559 additions and 576 deletions
@@ -0,0 +1,29 @@
+# Starter Dashboards
+
+Superset dashboard definitions for Stonks Oracle research, analysis, and trading review.
+
+## Dashboards
+- Symbol Overview — company profiles, source health, recent documents, and market snapshots
+- Sentiment Heatmap — market-wide sentiment by sector and symbol, catalyst analysis, contradiction tracking
+- Prediction Accuracy — predicted signals vs realized price moves, confidence calibration, per-symbol accuracy
+- Paper Trading PnL — cumulative PnL, daily performance, position snapshots, order history, and scorecards
+
+## Data Sources
+These dashboards query the Trino `lakehouse` catalog over MinIO-backed analytical fact tables:
+- `lakehouse.stonks.documents` — ingested document metadata
+- `lakehouse.stonks.document_extractions` — AI extraction outputs
+- `lakehouse.stonks.trade_signals` — aggregated trend signals
+- `lakehouse.stonks.market_bars` — OHLCV bar data
+- `lakehouse.stonks.prediction_vs_outcome` — prediction accuracy tracking
+- `lakehouse.stonks.pnl_daily` — daily PnL records
+- `lakehouse.stonks.positions_daily` — end-of-day position snapshots
+- `lakehouse.stonks.trade_orders` — order submission records
+- `lakehouse.stonks.trade_fills` — fill and execution records
+
+## Setup
+1. Import the dashboard JSON files into Superset via the Superset UI or CLI
+2. Ensure the Trino datasource is configured: `trino://trino@trino:8080/lakehouse/stonks`
+3. Create the lakehouse views from `lakehouse/views/` for additional drill-down capability
+
+## Trino Connection
+The dashboards use the default Superset Trino connection configured in `infra/superset/superset_config.py`.
@@ -0,0 +1,124 @@
+{
+  "dashboard_title": "Paper Trading PnL",
+  "description": "Paper trading performance tracking with PnL curves, position snapshots, order history, and trade detail drill-down.",
+  "slug": "paper-trading-pnl",
+  "position_json": {
+    "HEADER_ID": {"id": "HEADER_ID", "type": "HEADER", "meta": {"text": "Paper Trading PnL"}},
+    "ROW-1": {
+      "type": "ROW",
+      "children": ["CHART-total-net-pnl-kpi", "CHART-win-rate-kpi", "CHART-total-orders-kpi", "CHART-active-positions-kpi"]
+    },
+    "ROW-2": {
+      "type": "ROW",
+      "children": ["CHART-cumulative-pnl-timeseries", "CHART-daily-pnl-bar"]
+    },
+    "ROW-3": {
+      "type": "ROW",
+      "children": ["CHART-pnl-by-symbol", "CHART-order-status-pie"]
+    },
+    "ROW-4": {
+      "type": "ROW",
+      "children": ["CHART-positions-table"]
+    },
+    "ROW-5": {
+      "type": "ROW",
+      "children": ["CHART-scorecard-table"]
+    },
+    "ROW-6": {
+      "type": "ROW",
+      "children": ["CHART-recent-orders-table"]
+    }
+  },
+  "metadata": {
+    "refresh_frequency": 300,
+    "default_filters": "{}",
+    "color_scheme": "supersetColors"
+  },
+  "charts": [
+    {
+      "slice_name": "Total Net PnL",
+      "viz_type": "big_number_total",
+      "description": "Cumulative net PnL across all paper trading activity",
+      "datasource_type": "trino",
+      "query": "SELECT ROUND(SUM(net_pnl), 2) AS total_net_pnl FROM lakehouse.stonks.pnl_daily WHERE execution_mode = 'paper'"
+    },
+    {
+      "slice_name": "Win Rate",
+      "viz_type": "big_number_total",
+      "description": "Fraction of trading days with positive net PnL",
+      "datasource_type": "trino",
+      "query": "SELECT ROUND(CAST(COUNT(CASE WHEN net_pnl > 0 THEN 1 END) AS DOUBLE) / NULLIF(COUNT(*), 0), 4) AS win_rate FROM lakehouse.stonks.pnl_daily WHERE execution_mode = 'paper'"
+    },
+    {
+      "slice_name": "Total Orders",
+      "viz_type": "big_number_total",
+      "description": "Total paper trade orders submitted",
+      "datasource_type": "trino",
+      "query": "SELECT COUNT(DISTINCT order_id) AS total_orders FROM lakehouse.stonks.trade_orders WHERE execution_mode = 'paper'"
+    },
+    {
+      "slice_name": "Active Positions",
+      "viz_type": "big_number_total",
+      "description": "Number of symbols with open positions as of the latest snapshot",
+      "datasource_type": "trino",
+      "query": "SELECT COUNT(DISTINCT ticker) AS active_positions FROM lakehouse.stonks.positions_daily WHERE execution_mode = 'paper' AND quantity <> 0 AND dt = (SELECT MAX(dt) FROM lakehouse.stonks.positions_daily WHERE execution_mode = 'paper')"
+    },
+    {
+      "slice_name": "Cumulative PnL Over Time",
+      "viz_type": "echarts_timeseries_line",
+      "description": "Running cumulative net PnL across all paper trades",
+      "datasource_type": "trino",
+      "query": "SELECT dt AS bucket, SUM(net_pnl) AS daily_net_pnl, SUM(SUM(net_pnl)) OVER (ORDER BY dt) AS cumulative_pnl FROM lakehouse.stonks.pnl_daily WHERE execution_mode = 'paper' GROUP BY dt ORDER BY dt"
+    },
+    {
+      "slice_name": "Daily PnL",
+      "viz_type": "echarts_timeseries_bar",
+      "description": "Daily net PnL for paper trading, colored by positive/negative",
+      "datasource_type": "trino",
+      "query": "SELECT dt AS bucket, ROUND(SUM(net_pnl), 2) AS daily_pnl, ROUND(SUM(realized_pnl), 2) AS realized, ROUND(SUM(unrealized_pnl), 2) AS unrealized FROM lakehouse.stonks.pnl_daily WHERE execution_mode = 'paper' GROUP BY dt ORDER BY dt",
+      "params": {
+        "x_axis": "bucket",
+        "metrics": ["daily_pnl"]
+      }
+    },
+    {
+      "slice_name": "PnL by Symbol",
+      "viz_type": "echarts_timeseries_bar",
+      "description": "Total net PnL per symbol for paper trading",
+      "datasource_type": "trino",
+      "query": "SELECT ticker, ROUND(SUM(net_pnl), 2) AS total_pnl, ROUND(SUM(realized_pnl), 2) AS realized_pnl, ROUND(SUM(fees), 2) AS total_fees FROM lakehouse.stonks.pnl_daily WHERE execution_mode = 'paper' GROUP BY ticker ORDER BY total_pnl DESC",
+      "params": {
+        "x_axis": "ticker",
+        "metrics": ["total_pnl"]
+      }
+    },
+    {
+      "slice_name": "Order Status Distribution",
+      "viz_type": "pie",
+      "description": "Breakdown of paper trade order statuses",
+      "datasource_type": "trino",
+      "query": "SELECT status, COUNT(*) AS count FROM lakehouse.stonks.trade_orders WHERE execution_mode = 'paper' GROUP BY status ORDER BY count DESC"
+    },
+    {
+      "slice_name": "Current Positions",
+      "viz_type": "table",
+      "description": "Latest position snapshot for all paper trading symbols",
+      "datasource_type": "trino",
+      "query": "SELECT p.ticker, p.quantity, ROUND(p.avg_entry_price, 2) AS avg_entry, ROUND(p.close_price, 2) AS close_price, ROUND(p.market_value, 2) AS market_value, ROUND(p.unrealized_pnl, 2) AS unrealized_pnl, p.snapshot_at FROM lakehouse.stonks.positions_daily p WHERE p.execution_mode = 'paper' AND p.dt = (SELECT MAX(dt) FROM lakehouse.stonks.positions_daily WHERE execution_mode = 'paper') ORDER BY ABS(p.unrealized_pnl) DESC"
+    },
+    {
+      "slice_name": "Paper Trade Scorecard",
+      "viz_type": "table",
+      "description": "Per-symbol paper trading scorecard with win rates, PnL, and order counts",
+      "datasource_type": "trino",
+      "query": "SELECT pnl.ticker, COUNT(DISTINCT pnl.dt) AS trading_days, ROUND(SUM(pnl.net_pnl), 2) AS total_net_pnl, ROUND(AVG(pnl.net_pnl), 2) AS avg_daily_pnl, ROUND(CAST(COUNT(CASE WHEN pnl.net_pnl > 0 THEN 1 END) AS DOUBLE) / NULLIF(COUNT(*), 0), 4) AS win_rate, ROUND(MIN(pnl.net_pnl), 2) AS worst_day, ROUND(MAX(pnl.net_pnl), 2) AS best_day, ROUND(SUM(pnl.fees), 2) AS total_fees, MIN(pnl.dt) AS first_trade, MAX(pnl.dt) AS last_trade FROM lakehouse.stonks.pnl_daily pnl WHERE pnl.execution_mode = 'paper' GROUP BY pnl.ticker ORDER BY total_net_pnl DESC"
+    },
+    {
+      "slice_name": "Recent Orders",
+      "viz_type": "table",
+      "description": "Most recent paper trade orders with fill details",
+      "datasource_type": "trino",
+      "query": "SELECT o.ticker, o.side, o.order_type, o.quantity, ROUND(o.limit_price, 2) AS limit_price, o.status, f.fill_price, f.fill_quantity, f.commission, o.submitted_at, f.filled_at FROM lakehouse.stonks.trade_orders o LEFT JOIN lakehouse.stonks.trade_fills f ON o.order_id = f.order_id AND o.dt = f.dt WHERE o.execution_mode = 'paper' ORDER BY o.submitted_at DESC LIMIT 50"
+    }
+  ]
+}
@@ -0,0 +1,125 @@
+{
+  "dashboard_title": "Prediction Accuracy",
+  "description": "Predicted signals vs realized price moves, confidence calibration, and model accuracy tracking.",
+  "slug": "prediction-accuracy",
+  "position_json": {
+    "HEADER_ID": {"id": "HEADER_ID", "type": "HEADER", "meta": {"text": "Prediction Accuracy"}},
+    "ROW-1": {
+      "type": "ROW",
+      "children": ["CHART-overall-hit-rate-kpi", "CHART-total-predictions-kpi", "CHART-avg-confidence-kpi", "CHART-avg-move-kpi"]
+    },
+    "ROW-2": {
+      "type": "ROW",
+      "children": ["CHART-hit-rate-timeseries", "CHART-outcome-distribution-pie"]
+    },
+    "ROW-3": {
+      "type": "ROW",
+      "children": ["CHART-confidence-calibration", "CHART-confidence-vs-move-scatter"]
+    },
+    "ROW-4": {
+      "type": "ROW",
+      "children": ["CHART-accuracy-by-symbol", "CHART-accuracy-by-action"]
+    },
+    "ROW-5": {
+      "type": "ROW",
+      "children": ["CHART-recent-predictions-table"]
+    }
+  },
+  "metadata": {
+    "refresh_frequency": 600,
+    "default_filters": "{}",
+    "color_scheme": "supersetColors"
+  },
+  "charts": [
+    {
+      "slice_name": "Overall Hit Rate",
+      "viz_type": "big_number_total",
+      "description": "Fraction of predictions with correct directional outcome over the last 30 days",
+      "datasource_type": "trino",
+      "query": "SELECT ROUND(CAST(COUNT(CASE WHEN outcome = 'correct' THEN 1 END) AS DOUBLE) / NULLIF(COUNT(*), 0), 4) AS hit_rate FROM lakehouse.stonks.prediction_vs_outcome WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY"
+    },
+    {
+      "slice_name": "Total Predictions (30d)",
+      "viz_type": "big_number_total",
+      "description": "Total evaluated predictions in the last 30 days",
+      "datasource_type": "trino",
+      "query": "SELECT COUNT(*) AS total_predictions FROM lakehouse.stonks.prediction_vs_outcome WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY"
+    },
+    {
+      "slice_name": "Avg Predicted Confidence",
+      "viz_type": "big_number_total",
+      "description": "Average confidence of predictions in the last 30 days",
+      "datasource_type": "trino",
+      "query": "SELECT ROUND(AVG(predicted_confidence), 3) AS avg_confidence FROM lakehouse.stonks.prediction_vs_outcome WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY"
+    },
+    {
+      "slice_name": "Avg Realized Move",
+      "viz_type": "big_number_total",
+      "description": "Average absolute realized price move percentage",
+      "datasource_type": "trino",
+      "query": "SELECT ROUND(AVG(ABS(actual_move_pct)), 3) AS avg_abs_move FROM lakehouse.stonks.prediction_vs_outcome WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY"
+    },
+    {
+      "slice_name": "Daily Hit Rate",
+      "viz_type": "echarts_timeseries_line",
+      "description": "Daily prediction hit rate over the last 30 days",
+      "datasource_type": "trino",
+      "query": "SELECT dt AS bucket, COUNT(*) AS total, COUNT(CASE WHEN outcome = 'correct' THEN 1 END) AS correct, ROUND(CAST(COUNT(CASE WHEN outcome = 'correct' THEN 1 END) AS DOUBLE) / NULLIF(COUNT(*), 0), 4) AS hit_rate FROM lakehouse.stonks.prediction_vs_outcome WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY GROUP BY dt ORDER BY dt"
+    },
+    {
+      "slice_name": "Outcome Distribution",
+      "viz_type": "pie",
+      "description": "Breakdown of prediction outcomes (correct, incorrect, neutral) over the last 30 days",
+      "datasource_type": "trino",
+      "query": "SELECT outcome, COUNT(*) AS count FROM lakehouse.stonks.prediction_vs_outcome WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY GROUP BY outcome ORDER BY count DESC"
+    },
+    {
+      "slice_name": "Confidence Calibration",
+      "viz_type": "echarts_timeseries_bar",
+      "description": "Hit rate by confidence bucket to assess calibration quality",
+      "datasource_type": "trino",
+      "query": "SELECT CASE WHEN predicted_confidence >= 0.8 THEN '0.8-1.0 (high)' WHEN predicted_confidence >= 0.6 THEN '0.6-0.8 (medium)' WHEN predicted_confidence >= 0.4 THEN '0.4-0.6 (low)' ELSE '0.0-0.4 (very low)' END AS confidence_bucket, COUNT(*) AS total, COUNT(CASE WHEN outcome = 'correct' THEN 1 END) AS correct, ROUND(CAST(COUNT(CASE WHEN outcome = 'correct' THEN 1 END) AS DOUBLE) / NULLIF(COUNT(*), 0), 4) AS hit_rate FROM lakehouse.stonks.prediction_vs_outcome WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY GROUP BY 1 ORDER BY 1",
+      "params": {
+        "x_axis": "confidence_bucket",
+        "metrics": ["hit_rate"]
+      }
+    },
+    {
+      "slice_name": "Confidence vs Realized Move",
+      "viz_type": "echarts_timeseries_scatter",
+      "description": "Scatter plot of predicted confidence vs actual realized move percentage",
+      "datasource_type": "trino",
+      "query": "SELECT ticker, predicted_confidence, actual_move_pct, predicted_action, outcome, dt FROM lakehouse.stonks.prediction_vs_outcome WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY ORDER BY dt DESC",
+      "params": {
+        "x_axis": "predicted_confidence",
+        "y_axis": "actual_move_pct",
+        "groupby": ["outcome"]
+      }
+    },
+    {
+      "slice_name": "Accuracy by Symbol",
+      "viz_type": "table",
+      "description": "Per-symbol prediction accuracy summary",
+      "datasource_type": "trino",
+      "query": "SELECT ticker, COUNT(*) AS predictions, COUNT(CASE WHEN outcome = 'correct' THEN 1 END) AS correct, COUNT(CASE WHEN outcome = 'incorrect' THEN 1 END) AS incorrect, ROUND(CAST(COUNT(CASE WHEN outcome = 'correct' THEN 1 END) AS DOUBLE) / NULLIF(COUNT(*), 0), 4) AS hit_rate, ROUND(AVG(predicted_confidence), 3) AS avg_confidence, ROUND(AVG(actual_move_pct), 3) AS avg_move_pct, ROUND(AVG(ABS(actual_move_pct)), 3) AS avg_abs_move_pct FROM lakehouse.stonks.prediction_vs_outcome WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY GROUP BY ticker ORDER BY hit_rate DESC"
+    },
+    {
+      "slice_name": "Accuracy by Action Type",
+      "viz_type": "echarts_timeseries_bar",
+      "description": "Hit rate broken down by predicted action (buy, sell, hold, watch)",
+      "datasource_type": "trino",
+      "query": "SELECT predicted_action, COUNT(*) AS total, COUNT(CASE WHEN outcome = 'correct' THEN 1 END) AS correct, ROUND(CAST(COUNT(CASE WHEN outcome = 'correct' THEN 1 END) AS DOUBLE) / NULLIF(COUNT(*), 0), 4) AS hit_rate, ROUND(AVG(predicted_confidence), 3) AS avg_confidence FROM lakehouse.stonks.prediction_vs_outcome WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY GROUP BY predicted_action ORDER BY predicted_action",
+      "params": {
+        "x_axis": "predicted_action",
+        "metrics": ["hit_rate"]
+      }
+    },
+    {
+      "slice_name": "Recent Predictions",
+      "viz_type": "table",
+      "description": "Most recent evaluated predictions with outcomes",
+      "datasource_type": "trino",
+      "query": "SELECT ticker, predicted_action, ROUND(predicted_confidence, 3) AS confidence, ROUND(actual_move_pct, 3) AS actual_move_pct, outcome, horizon_days, model_version, predicted_at, evaluated_at FROM lakehouse.stonks.prediction_vs_outcome WHERE dt >= CURRENT_DATE - INTERVAL '14' DAY ORDER BY evaluated_at DESC LIMIT 50"
+    }
+  ]
+}
@@ -0,0 +1,120 @@
+{
+  "dashboard_title": "Sentiment Heatmap",
+  "description": "Market-wide sentiment visualization by sector and symbol, with trend direction and catalyst analysis.",
+  "slug": "sentiment-heatmap",
+  "position_json": {
+    "HEADER_ID": {"id": "HEADER_ID", "type": "HEADER", "meta": {"text": "Sentiment Heatmap"}},
+    "ROW-1": {
+      "type": "ROW",
+      "children": ["CHART-bullish-count-kpi", "CHART-bearish-count-kpi", "CHART-mixed-count-kpi", "CHART-avg-contradiction-kpi"]
+    },
+    "ROW-2": {
+      "type": "ROW",
+      "children": ["CHART-sentiment-heatmap"]
+    },
+    "ROW-3": {
+      "type": "ROW",
+      "children": ["CHART-sentiment-timeseries", "CHART-catalyst-breakdown"]
+    },
+    "ROW-4": {
+      "type": "ROW",
+      "children": ["CHART-contradiction-scatter", "CHART-sentiment-distribution"]
+    },
+    "ROW-5": {
+      "type": "ROW",
+      "children": ["CHART-symbol-sentiment-detail"]
+    }
+  },
+  "metadata": {
+    "refresh_frequency": 300,
+    "default_filters": "{}",
+    "color_scheme": "supersetColors"
+  },
+  "charts": [
+    {
+      "slice_name": "Bullish Signals (7d)",
+      "viz_type": "big_number_total",
+      "description": "Count of bullish trend signals in the last 7 days",
+      "datasource_type": "trino",
+      "query": "SELECT COUNT(*) AS bullish_count FROM lakehouse.stonks.trade_signals WHERE trend_direction = 'bullish' AND dt >= CURRENT_DATE - INTERVAL '7' DAY"
+    },
+    {
+      "slice_name": "Bearish Signals (7d)",
+      "viz_type": "big_number_total",
+      "description": "Count of bearish trend signals in the last 7 days",
+      "datasource_type": "trino",
+      "query": "SELECT COUNT(*) AS bearish_count FROM lakehouse.stonks.trade_signals WHERE trend_direction = 'bearish' AND dt >= CURRENT_DATE - INTERVAL '7' DAY"
+    },
+    {
+      "slice_name": "Mixed Signals (7d)",
+      "viz_type": "big_number_total",
+      "description": "Count of mixed or neutral trend signals in the last 7 days",
+      "datasource_type": "trino",
+      "query": "SELECT COUNT(*) AS mixed_count FROM lakehouse.stonks.trade_signals WHERE trend_direction IN ('mixed', 'neutral') AND dt >= CURRENT_DATE - INTERVAL '7' DAY"
+    },
+    {
+      "slice_name": "Avg Contradiction Score (7d)",
+      "viz_type": "big_number_total",
+      "description": "Average contradiction score across all signals in the last 7 days",
+      "datasource_type": "trino",
+      "query": "SELECT ROUND(AVG(contradiction_score), 3) AS avg_contradiction FROM lakehouse.stonks.trade_signals WHERE dt >= CURRENT_DATE - INTERVAL '7' DAY"
+    },
+    {
+      "slice_name": "Sentiment Heatmap by Symbol",
+      "viz_type": "heatmap",
+      "description": "Daily average sentiment impact score by symbol over the last 14 days",
+      "datasource_type": "trino",
+      "query": "SELECT de.ticker, de.dt, ROUND(AVG(de.impact_score), 3) AS avg_impact, AVG(CASE WHEN de.sentiment = 'positive' THEN 1.0 WHEN de.sentiment = 'negative' THEN -1.0 ELSE 0.0 END) AS sentiment_score FROM lakehouse.stonks.document_extractions de WHERE de.dt >= CURRENT_DATE - INTERVAL '14' DAY GROUP BY de.ticker, de.dt ORDER BY de.ticker, de.dt",
+      "params": {
+        "x_axis": "dt",
+        "y_axis": "ticker",
+        "metric": "sentiment_score"
+      }
+    },
+    {
+      "slice_name": "Sentiment Trend Over Time",
+      "viz_type": "echarts_timeseries_line",
+      "description": "Daily average sentiment score across all symbols over the last 30 days",
+      "datasource_type": "trino",
+      "query": "SELECT de.dt AS bucket, ROUND(AVG(CASE WHEN de.sentiment = 'positive' THEN 1.0 WHEN de.sentiment = 'negative' THEN -1.0 ELSE 0.0 END), 3) AS avg_sentiment, COUNT(*) AS extraction_count FROM lakehouse.stonks.document_extractions de WHERE de.dt >= CURRENT_DATE - INTERVAL '30' DAY GROUP BY de.dt ORDER BY de.dt"
+    },
+    {
+      "slice_name": "Catalyst Type Breakdown",
+      "viz_type": "pie",
+      "description": "Distribution of catalyst types across extractions in the last 14 days",
+      "datasource_type": "trino",
+      "query": "SELECT catalyst_type, COUNT(*) AS count FROM lakehouse.stonks.document_extractions WHERE dt >= CURRENT_DATE - INTERVAL '14' DAY AND catalyst_type IS NOT NULL GROUP BY catalyst_type ORDER BY count DESC"
+    },
+    {
+      "slice_name": "Contradiction vs Confidence",
+      "viz_type": "echarts_timeseries_scatter",
+      "description": "Scatter of contradiction score vs confidence for recent signals",
+      "datasource_type": "trino",
+      "query": "SELECT ticker, confidence, contradiction_score, trend_strength, trend_direction, dt FROM lakehouse.stonks.trade_signals WHERE dt >= CURRENT_DATE - INTERVAL '14' DAY ORDER BY dt DESC",
+      "params": {
+        "x_axis": "confidence",
+        "y_axis": "contradiction_score",
+        "groupby": ["trend_direction"]
+      }
+    },
+    {
+      "slice_name": "Sentiment Distribution by Symbol",
+      "viz_type": "echarts_timeseries_bar",
+      "description": "Count of positive, negative, and neutral extractions per symbol in the last 14 days",
+      "datasource_type": "trino",
+      "query": "SELECT ticker, sentiment, COUNT(*) AS count FROM lakehouse.stonks.document_extractions WHERE dt >= CURRENT_DATE - INTERVAL '14' DAY GROUP BY ticker, sentiment ORDER BY ticker, sentiment",
+      "params": {
+        "x_axis": "ticker",
+        "metrics": ["count"],
+        "groupby": ["sentiment"]
+      }
+    },
+    {
+      "slice_name": "Symbol Sentiment Detail",
+      "viz_type": "table",
+      "description": "Per-symbol sentiment summary with extraction counts, average impact, and dominant catalysts",
+      "datasource_type": "trino",
+      "query": "SELECT de.ticker, COUNT(*) AS extractions, ROUND(AVG(de.impact_score), 3) AS avg_impact, ROUND(AVG(de.confidence), 3) AS avg_confidence, ROUND(AVG(de.novelty_score), 3) AS avg_novelty, COUNT(CASE WHEN de.sentiment = 'positive' THEN 1 END) AS positive_count, COUNT(CASE WHEN de.sentiment = 'negative' THEN 1 END) AS negative_count, COUNT(CASE WHEN de.sentiment = 'neutral' THEN 1 END) AS neutral_count, ts.trend_direction AS latest_trend, ts.trend_strength AS latest_trend_strength FROM lakehouse.stonks.document_extractions de LEFT JOIN lakehouse.stonks.trade_signals ts ON de.ticker = ts.ticker AND ts.dt = (SELECT MAX(dt) FROM lakehouse.stonks.trade_signals WHERE ticker = de.ticker) WHERE de.dt >= CURRENT_DATE - INTERVAL '14' DAY GROUP BY de.ticker, ts.trend_direction, ts.trend_strength ORDER BY de.ticker"
+    }
+  ]
+}
@@ -0,0 +1,104 @@
+{
+  "dashboard_title": "Symbol Overview",
+  "description": "Company profiles, source health, recent documents, and market snapshot for tracked symbols.",
+  "slug": "symbol-overview",
+  "position_json": {
+    "HEADER_ID": {"id": "HEADER_ID", "type": "HEADER", "meta": {"text": "Symbol Overview"}},
+    "ROW-1": {
+      "type": "ROW",
+      "children": ["CHART-tracked-symbols-kpi", "CHART-total-documents-kpi", "CHART-total-extractions-kpi", "CHART-active-signals-kpi"]
+    },
+    "ROW-2": {
+      "type": "ROW",
+      "children": ["CHART-company-summary-table"]
+    },
+    "ROW-3": {
+      "type": "ROW",
+      "children": ["CHART-recent-documents-timeseries", "CHART-document-type-breakdown"]
+    },
+    "ROW-4": {
+      "type": "ROW",
+      "children": ["CHART-latest-prices-table"]
+    },
+    "ROW-5": {
+      "type": "ROW",
+      "children": ["CHART-recent-documents-table"]
+    }
+  },
+  "metadata": {
+    "refresh_frequency": 300,
+    "default_filters": "{}",
+    "color_scheme": "supersetColors"
+  },
+  "charts": [
+    {
+      "slice_name": "Tracked Symbols",
+      "viz_type": "big_number_total",
+      "description": "Count of distinct symbols with documents in the last 30 days",
+      "datasource_type": "trino",
+      "query": "SELECT COUNT(DISTINCT ticker) AS tracked_symbols FROM lakehouse.stonks.documents WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY"
+    },
+    {
+      "slice_name": "Total Documents (30d)",
+      "viz_type": "big_number_total",
+      "description": "Total documents ingested in the last 30 days",
+      "datasource_type": "trino",
+      "query": "SELECT COUNT(*) AS total_documents FROM lakehouse.stonks.documents WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY"
+    },
+    {
+      "slice_name": "Total Extractions (30d)",
+      "viz_type": "big_number_total",
+      "description": "Total AI extractions completed in the last 30 days",
+      "datasource_type": "trino",
+      "query": "SELECT COUNT(*) AS total_extractions FROM lakehouse.stonks.document_extractions WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY"
+    },
+    {
+      "slice_name": "Active Signals (7d)",
+      "viz_type": "big_number_total",
+      "description": "Trade signals generated in the last 7 days",
+      "datasource_type": "trino",
+      "query": "SELECT COUNT(*) AS active_signals FROM lakehouse.stonks.trade_signals WHERE dt >= CURRENT_DATE - INTERVAL '7' DAY"
+    },
+    {
+      "slice_name": "Company Summary",
+      "viz_type": "table",
+      "description": "Per-symbol summary with document counts, extraction counts, latest signal, and latest price",
+      "datasource_type": "trino",
+      "query": "SELECT d.ticker, COUNT(DISTINCT d.document_id) AS documents_30d, COUNT(DISTINCT de.document_id) AS extractions_30d, MAX(d.published_at) AS latest_document_at, MAX(ts.generated_at) AS latest_signal_at, MAX(ts.trend_direction) AS latest_trend, MAX(mb.close_price) AS latest_close FROM lakehouse.stonks.documents d LEFT JOIN lakehouse.stonks.document_extractions de ON d.ticker = de.ticker AND de.dt >= CURRENT_DATE - INTERVAL '30' DAY LEFT JOIN lakehouse.stonks.trade_signals ts ON d.ticker = ts.ticker AND ts.dt = (SELECT MAX(dt) FROM lakehouse.stonks.trade_signals WHERE ticker = d.ticker) LEFT JOIN lakehouse.stonks.market_bars mb ON d.ticker = mb.ticker AND mb.dt = (SELECT MAX(dt) FROM lakehouse.stonks.market_bars WHERE ticker = d.ticker) WHERE d.dt >= CURRENT_DATE - INTERVAL '30' DAY GROUP BY d.ticker ORDER BY d.ticker"
+    },
+    {
+      "slice_name": "Documents Ingested Over Time",
+      "viz_type": "echarts_timeseries_bar",
+      "description": "Daily document ingestion counts by source type over the last 30 days",
+      "datasource_type": "trino",
+      "query": "SELECT dt AS bucket, source_type, COUNT(*) AS doc_count FROM lakehouse.stonks.documents WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY GROUP BY dt, source_type ORDER BY dt",
+      "params": {
+        "x_axis": "bucket",
+        "metrics": ["doc_count"],
+        "groupby": ["source_type"],
+        "time_grain_sqla": "P1D"
+      }
+    },
+    {
+      "slice_name": "Document Type Breakdown",
+      "viz_type": "pie",
+      "description": "Distribution of documents by type in the last 30 days",
+      "datasource_type": "trino",
+      "query": "SELECT document_type, COUNT(*) AS count FROM lakehouse.stonks.documents WHERE dt >= CURRENT_DATE - INTERVAL '30' DAY GROUP BY document_type ORDER BY count DESC"
+    },
+    {
+      "slice_name": "Latest Prices by Symbol",
+      "viz_type": "table",
+      "description": "Most recent closing prices and volume for each tracked symbol",
+      "datasource_type": "trino",
+      "query": "SELECT mb.ticker, mb.close_price, mb.open_price, mb.high_price, mb.low_price, mb.volume, mb.vwap, mb.bar_timestamp FROM lakehouse.stonks.market_bars mb INNER JOIN (SELECT ticker, MAX(bar_timestamp) AS max_ts FROM lakehouse.stonks.market_bars GROUP BY ticker) latest ON mb.ticker = latest.ticker AND mb.bar_timestamp = latest.max_ts ORDER BY mb.ticker"
+    },
+    {
+      "slice_name": "Recent Documents",
+      "viz_type": "table",
+      "description": "Most recently ingested documents across all symbols",
+      "datasource_type": "trino",
+      "query": "SELECT ticker, document_type, source_type, title, publisher, published_at, retrieved_at, confidence FROM lakehouse.stonks.documents WHERE dt >= CURRENT_DATE - INTERVAL '7' DAY ORDER BY retrieved_at DESC LIMIT 50"
+    }
+  ]
+}