| name | databricks-aibi-dashboards |
| description | Create Databricks AI/BI dashboards. Must use when creating, updating, or deploying Lakeview dashboards as Databricks Dashboard have a unique json structure. CRITICAL: You MUST test ALL SQL queries via CLI BEFORE deploying. Follow guidelines strictly. |
| compatibility | Requires databricks CLI (>= v1.0.0) |
| metadata | {"version":"0.2.0"} |
| parent | databricks-core |
AI/BI Dashboard Skill
Create Databricks AI/BI dashboards (formerly Lakeview dashboards).
A dashboard should be showing something relevant for a human, typically some KPI on the top, and based on the story, some graph (often temporal), and we see "something happens".
Follow these guidelines strictly.
When a custom app fits better: A managed AI/BI dashboard is the right tool for read-only KPIs, charts, and filters over governed tables. If the user instead needs a custom-code interactive app — write-back / data entry, bespoke UI or interactions beyond the dashboard grid, embedded or auth-gated workflows, or a conversational Genie/chat assistant as the primary surface — build a Databricks App instead with the databricks-apps skill (which brings in databricks-app-design for the data-screen UX). Linking an "Ask Genie" space to this dashboard stays here (see Linking a Genie Space below).
Quick Reference
| Task | Command |
|---|
| List warehouses | databricks warehouses list |
| List tables | databricks experimental aitools tools query --warehouse WH "SHOW TABLES IN catalog.schema" |
| Get schema | databricks experimental aitools tools discover-schema catalog.schema.table1 catalog.schema.table2 |
| Test query | databricks experimental aitools tools query --warehouse WH "SELECT..." |
| Create dashboard | databricks lakeview create --display-name "X" --warehouse-id "WH" --dataset-catalog CATALOG --dataset-schema SCHEMA --serialized-dashboard "$(cat file.json)" --json '{"parent_path": "/Workspace/Users/<you>/path"}' — --dataset-catalog / --dataset-schema are flag-only (REQUIRED; CLI silently drops them if put in --json); parent_path is JSON-only (no flag). Queries must use bare table names. |
| Update dashboard | databricks lakeview update DASHBOARD_ID --serialized-dashboard "$(cat file.json)" |
| Publish | databricks lakeview publish DASHBOARD_ID --warehouse-id WH |
| Delete | databricks lakeview trash DASHBOARD_ID |
--warehouse flag: if databricks experimental aitools tools query --warehouse WH "..." fails with unknown flag: --warehouse on your CLI version, set DATABRICKS_WAREHOUSE_ID=WH in the environment instead and drop the flag — the command auto-picks it from there.
Widget Index (Version + Where Documented)
Wrong version = broken widget! This is the #1 cause of dashboard errors.
Cohort retention charts are built as a pivot with a color-scale cell style — there is no cohort widget type. See pivot in 2-advanced-widget-specifications.md.
NEW DASHBOARD CREATION WORKFLOW
You MUST test ALL SQL queries via CLI BEFORE deploying. Follow the overall logic in these steps for new dashboard - Skipping validation causes broken dashboards.
Step 1: Get Warehouse ID if not already known
databricks warehouses list
Step 2: Discover Table Schemas and existing data pattern
A good dashboard comes from knowing the data first. Spend time here — the exploration drives design decisions in Step 4 (which widgets, which filters, which groupings).
Use discover-schema as the default — one call returns columns, types, sample rows, null counts, and row count. If you only know the schema, list tables first with query "SHOW TABLES IN ...".
databricks experimental aitools tools discover-schema catalog.schema.orders catalog.schema.customers
Sample rows alone don't tell you what to build. you can write aggregate SQL through databricks experimental aitools tools query --warehouse <WH> "..." to probe typically:
- Cardinality of candidate grouping columns → decides chart color-group vs. table (≤8 distinct values for charts, see Cardinality & Readability below).
- Top categorical values → populates filter options and chart legends meaningfully.
- Numeric distribution (min/max/avg/percentiles) → decides KPI with delta vs. trend chart (flat metrics shouldn't be line charts, see Data Variance Considerations below).
- Trend viability at daily/weekly/monthly grain → picks the right trend granularity.
- Story confirmation — run the aggregations you plan to put in the dashboard and check they're not flat, empty, or uninteresting. Fix the query or adjust the story before moving on.
Fan out independent probes in one call — pass several positional SQLs (and/or repeated --file) and they run in parallel (default --concurrency 8):
DATABRICKS_WAREHOUSE_ID=<WH> databricks experimental aitools tools query --output json \
"SELECT COUNT(*) FROM catalog.schema.orders" \
"SELECT region, COUNT(*) FROM catalog.schema.orders GROUP BY region ORDER BY 2 DESC LIMIT 10" \
"SELECT MIN(ts), MAX(ts) FROM catalog.schema.orders"
--output json is mandatory in multi-query mode. Returns one object per statement: {sql, state, rows, error}; failures are per-statement (state: "FAILED"), others still succeed.- ⚠️ Don't trust the exit code (a failed statement can still exit
0) — gate on each object's state != "SUCCEEDED".
Dashboard queries are different — inside the dashboard JSON, the FROM clause must reference ONLY the table name, with no catalog or schema prefix:
- ✅ Correct:
FROM trips
- ❌ Wrong:
FROM nyctaxi.trips
- ❌ Wrong:
FROM samples.nyctaxi.trips
The catalog and schema are supplied separately via the --dataset-catalog and --dataset-schema flags when you run databricks lakeview create. These flags do NOT rewrite the query — they only fill in the catalog/schema when the query omits them. If you hardcode a catalog or schema in the FROM clause, the flags are ignored for that query and the dashboard won't be portable across environments.
Step 3: Verify Data Matches Story
The datasets.querylines in the dashboard json (see example below) must be tested to ensure
Before finalizing, run the SQL Queries you intend to add in each dataset to confirm that they run properly and that the result are valid.
This is crucial, as the widget defined in the json will use the query field output to render the visualization. The value should also make sense at a business level.
Remember that for the filter to work, the query should have the field available (so typically group by the filter field)
If values don't match expectations, ensure the query is correct, fix the data if you can, or adjust the story before creating the dashboard.
Step 4: Plan Dashboard Structure
Before writing JSON, plan your dashboard:
- You must know the expected specific JSON structure. For this, Read reference files: 1-widget-specifications.md, 3-filters.md.
Always make sure you read an entire example to understand the structure, like 4-examples.md.
-
Think: What widgets? Map each visualization to a dataset:
| Widget | Type | Dataset | Has filter field? |
|---|
| Revenue KPI | counter | ds_sales | ✓ date, region |
| Trend Chart | line | ds_sales | ✓ date, region |
| Top Products | table | ds_products | ✗ no date |
| ... | | | |
-
What filters? For each filter, verify ALL datasets you want filtered contain the filter field.
Filters only affect datasets that have the filter field. A pre-aggregated table without dates WON'T be date-filtered.
-
Build the dashboard JSON as a local working file (intermediate step, not the deliverable).
Step 5: Deploy
Now deploy the JSON to the workspace. Run databricks lakeview create (below). Your task is not complete until this command succeeds and returns a dashboard ID — the JSON file alone is an intermediate working artifact.
After deploying, the same lakeview subcommands manage the dashboard's lifecycle (list, get, update, publish, trash).
databricks lakeview create \
--display-name "My Dashboard" \
--warehouse-id "abc123def456" \
--dataset-catalog "my_catalog" \
--dataset-schema "my_schema" \
--serialized-dashboard "$(cat dashboard.json)" \
--json '{"parent_path": "/Workspace/Users/me@co.com/dashboards"}'
databricks lakeview list
databricks lakeview get DASHBOARD_ID
databricks lakeview update DASHBOARD_ID --serialized-dashboard "$(cat dashboard.json)"
databricks lakeview publish DASHBOARD_ID --warehouse-id WAREHOUSE_ID
databricks lakeview unpublish DASHBOARD_ID
databricks lakeview trash DASHBOARD_ID
databricks workspace-entity-tag-assignments create-tag-assignment \
dashboards DASHBOARD_ID aidevkit_project --tag-value ai-dev-kit
JSON Structure (Required Skeleton)
Every dashboard's serialized_dashboard content must follow this exact structure:
Important: ALWAYS add a space or \n at the end of each queryLines value as they are concatenated to create the dataset.
{
"datasets": [
{
"name": "ds_x",
"displayName": "Dataset X",
"queryLines": ["SELECT col1, col2 ", "FROM my_table"]
}
],
"pages": [
{
"name": "main",
"displayName": "Main",
"pageType": "PAGE_TYPE_CANVAS",
"layout": [
{"widget": {}, "position": {"x":0,"y":0,"width":2,"height":3}}
]
}
]
}
Structural rules (violations cause "failed to parse serialized dashboard"):
queryLines: Array of strings, NOT "query": "string". Elements are joined verbatim with no separator — end each line with or \n (or strip -- comments). A line ending in -- comment with no newline swallows the next line.- Widgets: INLINE in
layout[].widget, NOT a separate "widgets" array
pageType: Required on every page (PAGE_TYPE_CANVAS or PAGE_TYPE_GLOBAL_FILTERS)- Query binding:
query.fields[].name must exactly match encodings.*.fieldName
Theme & Color (always set this — it makes or breaks the dashboard)
Top-level uiSettings.theme controls colors, fonts, and widget chrome across every widget on the dashboard. Without it, the dashboard inherits the workspace default and looks generic. Set the full block on every dashboard you create — a coherent palette is the single highest-impact polish item.
Mental model — 60/30/10 rule mapped to theme keys: 60% neutral = canvas/widget/border backgrounds (set widgetBorderColor = widgetBackgroundColor to hide borders); 30% secondary = fontColor + visualizationColors (the content weight); 10% accent = selectionColor for filters / tabs / active selections — pick something distinct from text and palette; a safe-blue around #2272B4 matches the hyperlink convention and works as a default.
{
"datasets": [...],
"pages": [...],
"uiSettings": {
"theme": {
"canvasBackgroundColor": {"light": "#FCFCFC", "dark": "#1F272D"},
"widgetBackgroundColor": {"light": "#FFFFFF", "dark": "#11171C"},
"fontColor": {"light": "#11171C", "dark": "#E8ECF0"},
"selectionColor": {"light": "#2272B4", "dark": "#8ACAFF"},
"visualizationColors": [
"#FFA600", "#FF7054", "#DE5582", "#995495",
"#4E5185", "#1D425C", "#99DDB4"
],
"widgetHeaderAlignment": "LEFT"
}
}
}
Theme keys (mechanics):
visualizationColors: ordered palette every chart series and category mapping cycles through. Positions are 0-indexed: position: 0 = first color (#FFA600 above), position: 6 = seventh (#99DDB4). Length 5–8 is typical.- Background / font / selection colors take
light + dark pairs; the dashboard auto-selects based on viewer mode.
widgetHeaderAlignment: "LEFT" (default), "CENTER", or "RIGHT".- Per-widget color references:
{"themeColorType": "visualizationColors", "position": N} (0-indexed) to pin to a palette slot, or {"hex": "#FF0000"} for an exact color outside the palette.
Palette-design rules (this is what separates a polished dashboard from a noisy one):
- One coherent color family per dashboard, distinct across the suite. Walk across hues (e.g., amber → coral → pink → purple → navy), not one color faded toward white — a single-hue lightness ramp reads as one color and the viewer can't tell categories apart. Adjacent stops must be visually distinct: if you squint and two blur into one, push them further apart. Single-hue ramps are for quantitative widgets only (
colorRamp.mode: "custom-sequential"), never for visualizationColors.
- Pin semantic colors as literal hex, outside the palette. "Bad" = a warm coral (e.g.
#FF7E5C), "good" = a calm teal/green. Use color.scale.mappings with a bare hex string — {"value": "Critical", "color": "#FF7E5C"} — not {"hex": "..."} or themeColorType: position (both are silently dropped on chart widgets). Reuse the good-teal that's already in the palette so it never clashes.
- Color non-categorical widgets explicitly so they join the family. Maps & heatmaps:
colorRamp.mode: "custom-sequential" with {start, end} from the family (if directional: start = bad color, end = good color). Forecast / multi-series: pin per-series via color.scale.mappings keyed on displayName (actual = solid family color, forecast = contrast/alert, threshold = muted tone). Sparkline counters: set value.color to a family color, not grey.
- "Lighter / more pastel" tweak: nudge all stops up in lightness together; don't recolor individual ones. Re-sync the pinned semantic hex values; keep enough contrast on the alert color that it still reads as a warning.
Starter palettes (pick one and adapt — extend to 7-8 stops if needed; semantic red/green stay as literal hex per rule 2):
#094074 #3C6997 #5ADBFF #FFDD4A #FE9000
#003F5C #594E90 #BC4C96 #FF5F66 #FFA600
#4A8CC7 #F59770 #FFD84A #F0E09E #6DD980
#440154 #3B528B #21918C #5EC962 #FDE725
#4E79A7 #F28E2C #E15759 #76B7B2 #59A14F
#0072B2 #E69F00 #009E73 #CC79A7 #D55E00
#0D0887 #7E03A8 #CC4778 #F89441 #F0F921
#6929C4 #1192E8 #005D5D #9F1853 #FA4D56
~4-5% of viewers have color blindness (mostly red/green). Rows 4 and 6 above (viridis, Okabe-Ito) are CB-safe by design; verify customized palettes via simulator (Adobe Color, colorbrewer2.org). Don't put red and green adjacent, and rely on lightness contrast — not hue alone — between adjacent stops.
Linking a Genie Space (Optional)
To add an "Ask Genie" button to the dashboard, or to link a genie space/room with an ID, add uiSettings.genieSpace to the JSON (alongside theme if you have one):
"uiSettings": {
"theme": { },
"genieSpace": {
"isEnabled": true,
"overrideId": "your-genie-space-id-here",
"enablementMode": "ENABLED"
}
}
Genie is NOT a widget. Link via uiSettings.genieSpace only. There is no "widgetType": "assistant".
Design Best Practices
Apply unless user specifies otherwise:
- Global date filter: When data has temporal columns, add a date range filter. Most dashboards need time-based filtering.
- KPI time bounds: Use time-bounded metrics that enable period comparison (MoM, YoY). Unbounded "all-time" totals are less actionable.
- Value formatting: Format values based on their meaning — currency with symbol, percentages with %, large numbers compacted (K/M/B).
- Chart selection: Match cardinality to chart type. Few distinct values → bar with color grouping (or pie if you really want a snapshot); many values → table.
Reference Files
Before generating any dashboard JSON, read 4-examples.md first. It's a complete reference dashboard exercising every construct (dataset measures + MEASURE(), sparkline counters, forecast-line with annotations, pivot with conditional cells, symbol-map, histogram, range-slider filter, theme). Use it to learn the JSON shape; then adapt to the user's data and demo story — keep the structure, swap the tables, metrics, palette, and narrative for the case you're building.
Implementation Guidelines
1) DATASET ARCHITECTURE
- Fewer datasets is better — aim for one dataset that backs as many widgets as possible. Clicking a value on a chart (e.g., a bar, a slice) acts as a filter on that dataset, and every other widget sharing the same dataset re-renders with the click applied. Splitting widgets across many narrow datasets breaks this cross-filtering and forces users to set explicit filter widgets for what should "just work". Prefer one wide dataset per domain (orders, cases, customers); only split when a widget genuinely needs different grain, pre-aggregation, or a parameter the others can't tolerate.
- Two ways to define a dataset:
- SQL query:
{"name": "ds_x", "displayName": "...", "queryLines": ["SELECT ...", "FROM table"]} — full control, can include WITH / JOIN / AI_FORECAST / etc.
- UC asset shorthand:
{"name": "ds_x", "displayName": "...", "asset_name": "catalog.schema.table_or_view"} — no SQL needed. Works for regular tables, views, and metric views.
- Exactly ONE valid SQL query per dataset when using
queryLines (no multiple queries separated by ;)
- Queries must use bare table names only — no catalog, no schema prefix. Example:
FROM orders, never FROM gold.orders or FROM main.gold.orders. The catalog and schema come from the --dataset-catalog and --dataset-schema flags at creation time. These flags only fill in missing parts — they do NOT override any catalog/schema written in the query.
- SELECT must include all dimensions needed by widgets and all derived columns via
AS aliases
- Put ALL business logic (CASE/WHEN, COALESCE, ratios) into the dataset SELECT with explicit aliases
- Contract rule: Every widget
fieldName must exactly match a dataset column or alias
- Add ORDER BY when visualization depends on data order:
- Time series:
ORDER BY date for chronological display
- Rankings/Top-N:
ORDER BY metric DESC LIMIT 10 for "Top 10" charts
- Categorical charts:
ORDER BY metric DESC to show largest values first
Dataset-level measures + MEASURE()
Widget expressions are usually inline aggregations ({"name": "sum(x)", "expression": "SUM(\x`)"}`). But you can also declare reusable measures on the dataset itself and reference them by name — every widget that consumes the dataset can use the same metric without redefining it.
Two ways to define measures:
-
Dashboard-level columns (works on any dataset — SQL query or asset_name):
{
"name": "ds_support",
"queryLines": ["SELECT * FROM support_cases"],
"columns": [
{"displayName": "Total Cases", "description": "Count of cases",
"expression": "COUNT(`case_id`)"},
{"displayName": "Reopen Rate %", "description": "% of reopened cases",
"expression": "SUM(CASE WHEN `reopened_flag` THEN 1 ELSE 0 END) * 100.0 / COUNT(`case_id`)"},
{"displayName": "Priority Level", "description": "Sorted priority label",
"expression": "CASE WHEN `priority`='Critical' THEN '1-Critical' ELSE '4-Low' END"}
]
}
-
Metric-view source — if the dataset's asset_name (or FROM clause) is a UC metric view, its YAML-defined measures are already queryable. Do not redeclare them in columns. See databricks-metric-views.
Either way, widgets reference the measure by name:
"fields": [{"name": "measure(Total Cases)", "expression": "MEASURE(`Total Cases`)"}],
"encodings": {"value": {"fieldName": "measure(Total Cases)", "displayName": "Total Cases"}}
MEASURE(\...`)` works in counter, table, bar, line, pie, pivot — any widget that takes a field expression. Mix it with inline aggregations freely.
2) WIDGET FIELD EXPRESSIONS
CRITICAL: Field Name Matching Rule
The name in query.fields MUST exactly match the fieldName in encodings.
If they don't match, the widget shows "no selected fields to visualize" error!
Correct pattern for aggregations:
{"name": "sum(spend)", "expression": "SUM(`spend`)"}
{"fieldName": "sum(spend)", "displayName": "Total Spend"}
WRONG - names don't match:
{"name": "spend", "expression": "SUM(`spend`)"}
{"fieldName": "sum(spend)", ...}
Allowed expressions in widget queries (you CANNOT use CAST or other SQL in expressions):
{"name": "(sum|avg|count|countdistinct|min|max)(col)", "expression": "(SUM|AVG|COUNT|COUNT(DISTINCT)|MIN|MAX)(`col`)"}
{"name": "(daily|weekly|monthly)(date)", "expression": "DATE_TRUNC(\"(DAY|WEEK|MONTH)\", `date`)"}
{"name": "field", "expression": "`field`"}
If you need conditional logic or multi-field formulas, compute a derived column in the dataset SQL first.
3) SPARK SQL PATTERNS
- Date math:
date_sub(current_date(), N) for days, add_months(current_date(), -N) for months
- Date truncation:
DATE_TRUNC('DAY'|'WEEK'|'MONTH'|'QUARTER'|'YEAR', column)
- AVOID
INTERVAL syntax - use functions instead
4) LAYOUT (12-Column Grid, NO GAPS)
Every page must include "layoutVersion": "GRID_V1" alongside pageType.
{
"name": "overview",
"displayName": "Overview",
"pageType": "PAGE_TYPE_CANVAS",
"layoutVersion": "GRID_V1",
"layout": [...]
}
Each widget has a position: {"x": 0, "y": 0, "width": 4, "height": 4}
Pick the subdivision based on the audience. The 12-column grid divides cleanly into 3, 4, or 6 columns: a 3-column layout (each widget width: 4) reduces cognitive load and fits an executive overview; a 4-column (width: 3) is the all-rounder; a 6-column (width: 2) packs the most density for technical / operations dashboards where the reader is hunting through many metrics at once.
Default rule: each row should fill width=12 exactly — no gaps. Once you're confident with the grid, you can stagger heights across columns (a tall widget on the left paired with several shorter ones on the right) so the two halves don't share row boundaries — see 4-examples.md for the pattern. Start with strict rows; relax only when the stagger reads better visually.
CORRECT: WRONG:
y=0: [w=12] y=0: [w=8]____ ← gap!
y=1: [w=4][w=4][w=4] ← fills 12 y=1: [w=2][w=2][w=2][w=2]__ ← gap!
y=4: [w=6][w=6] ← fills 12
Recommended widget sizes:
| Widget Type | Width | Height | Notes |
|---|
| Text header | 12 | 1 | Full width; use SEPARATE widgets for title and subtitle |
| Counter/KPI | 4 | 3-4 | NEVER height=2 - too cramped! |
| Line/Bar/Area chart | 6 | 5-6 | Pair side-by-side to fill row |
| Pie chart | 6 | 5-6 | Needs space for legend |
| Full-width chart | 12 | 5-7 | For detailed time series |
| Table | 12 | 5-8 | Full width for readability |
Standard dashboard structure:
y=0: Title (w=12, h=1) - Dashboard title (use separate widget!)
y=1: Subtitle (w=12, h=1) - Description (use separate widget!)
y=2: KPIs (w=4 each, h=3) - 3 key metrics side-by-side
y=5: Section header (w=12, h=1) - "Trends" or similar
y=6: Charts (w=6 each, h=5) - Two charts side-by-side
y=11: Section header (w=12, h=1) - "Details"
y=12: Table (w=12, h=6) - Detailed data
5) CARDINALITY & READABILITY (CRITICAL)
Dashboard readability depends on limiting distinct values:
| Dimension Type | Max Values | Examples |
|---|
| Chart color/groups | 3-8 | 4 regions, 5 product lines, 3 tiers |
| Filters | 4-15 | 8 countries, 5 channels |
| High cardinality | Table only | customer_id, order_id, SKU |
Before creating any chart with color/grouping:
- Check column cardinality via discover-schema or a COUNT DISTINCT query
- If >10 distinct values, aggregate to higher level OR use TOP-N + "Other" bucket
- For high-cardinality dimensions, use a table widget instead of a chart
6) QUALITY CHECKLIST
Before deploying, verify:
- All widget names use only alphanumeric + hyphens + underscores
- Every page has
"layoutVersion": "GRID_V1"
- All rows sum to width=12 with no gaps
- KPIs use height 3-4, charts use height 5-6
- Chart dimensions have reasonable cardinality (≤8 for colors/groups)
- All widget fieldNames match dataset columns exactly
- Field
name in query.fields matches fieldName in encodings exactly (e.g., both "sum(spend)")
- Counter datasets: use
disaggregated: true for 1-row datasets, disaggregated: false with aggregation for multi-row
- Percent values must be 0-1 for
number-percent format (0.865 displays as "86.5%", don't forget to set the format). If data is 0-100, either divide by 100 in SQL or use number format instead.
- SQL uses Spark syntax (date_sub, not INTERVAL)
- All SQL queries tested via CLI and return expected data
- Every dataset you want filtered MUST contain the filter field — filters only affect datasets with that column in their query
Data Variance Considerations
Before creating trend charts, check if the metric has enough variance to visualize meaningfully:
SELECT MIN(metric), MAX(metric), MAX(metric) - MIN(metric) as range FROM dataset
If the range is very small relative to the scale (e.g., 83-89% on a 0-100 scale), the chart will appear nearly flat. Consider:
- Showing as KPI with delta/comparison instead of chart
- Using a table to display exact values
- Adjusting the visualization to focus on the variance
Related Skills
