schemaVersion: 3
meta:
  title: LiteLLM team cost board (demo)
  description: Version 2.0.0. Synthetic team cost dashboard. No credentials or database connection.
  codeLanguage: PYTHON
  timezone: UTC
  autoRerunApp: true
  logicQueryCacheTimeout:
    type: disabled
  publishedQueryCacheTimeout:
    type: disabled
  appTheme: LIGHT
cells:
- cellId: 9f2ee2c1-c9b7-5e03-853b-51ba6c25f7dc
  cellLabel: Read me
  cellType: MARKDOWN
  config:
    source: '# LiteLLM team cost board

      Template v2.0.0 · Synthetic demo. Filter by team and compare daily spend, unit cost and model mix.
      Use the Postgres template for your data.

      '
- cellId: 3989fc93-d6e3-5907-9e0d-e19b81b48b76
  cellLabel: Settings
  cellType: CODE
  config:
    source: "import pandas as pd\nimport numpy as np\nimport plotly.express as px\nfrom datetime import\
      \ datetime, timezone\n\nTEMPLATE_VERSION = '2.0.0'\npx.defaults.template = 'plotly_white'\npx.defaults.color_discrete_sequence\
      \ = ['#263cdb', '#0f766e', '#b45309', '#64748b']\n# UTC dates: end is exclusive. Leave blank to\
      \ use the last 30 complete days.\nreport_end = pd.Timestamp(datetime.now(timezone.utc).date())\n\
      report_start = report_end - pd.Timedelta(days=30)\n# Add one row per team for this exact report\
      \ window after Finance validates it.\n# value_usd: realized revenue or approved avoided cost; other_cost_usd:\
      \ infra,\n# automation, licenses and labor. Avoid counting the same benefit twice.\nvalue_data =\
      \ pd.DataFrame(columns=['team_id', 'value_usd', 'other_cost_usd',\n                            \
      \       'coverage_start', 'coverage_end'])\n"
- cellId: bcd197b2-1412-5f82-8de6-abe118fb7c81
  cellLabel: Source
  cellType: CODE
  config:
    source: "import pandas as pd\nfrom datetime import datetime, timezone, timedelta\n\n# Template v2.0.0.\
      \ Synthetic data for a reproducible preview.\n_demo_end = datetime.now(timezone.utc).date()\n_demo_rows\
      \ = []\nfor _n in range(60):\n    _day = _demo_end - timedelta(days=60-_n)\n    for _i, _team in\
      \ enumerate(['support-assistant', 'document-review', 'internal-search']):\n        _requests = 900\
      \ + _i * 300 + (_n % 7) * 40\n        _rate = [0.006, 0.012, 0.003][_i]\n        if _team == 'document-review'\
      \ and _n >= 53:\n            _rate *= 1.8\n        _seconds = [2.1, 4.2, 1.3][_i]\n        _demo_rows.append(dict(\n\
      \            day=_day, team_id=['support', 'legal', 'engineering'][_i], team_name=['Support assistant',\
      \ 'Document review', 'Internal search'][_i],\n            model='demo-premium-v2' if _team == 'document-review'\
      \ and _n >= 53 else 'demo-standard',\n            requests=_requests, spend_usd=round(_requests*_rate,\
      \ 4), unknown_cost_requests=0,\n            tokens=_requests*650, failures=_requests//100, unknown_status_requests=0,\n\
      \            response_cache_hits=_requests//10, response_cache_known=_requests,\n            latency_seconds_sum=_requests*_seconds,\
      \ latency_count=_requests,\n            ttft_seconds_sum=_requests*0.7, ttft_count=_requests))\n\
      spend_daily = pd.DataFrame(_demo_rows)\n"
- cellId: 8c3aac23-e134-5ec1-b0fa-1116428dc185
  cellLabel: Team ID (blank = all)
  cellType: INPUT
  config:
    name: team_filter
    outputType: STRING
    inputType: TEXT_INPUT
    defaultValue: ''
    options:
      multiline: false
- cellId: 777c6579-f9fe-5a12-afb5-ec5dcaae10c6
  cellLabel: Team name (blank = all)
  cellType: INPUT
  config:
    name: team_name_filter
    outputType: STRING
    inputType: TEXT_INPUT
    defaultValue: ''
    options:
      multiline: false
- cellId: 4c4a2d1d-9655-5ad6-8b39-b2360056d560
  cellLabel: Calculate board
  cellType: CODE
  config:
    source: "data = spend_daily.copy()\ndata['day'] = pd.to_datetime(data['day']).dt.normalize()\n_numeric\
      \ = [c for c in data.columns if c not in ['day', 'team_id', 'team_name', 'model']]\ndata[_numeric]\
      \ = data[_numeric].apply(pd.to_numeric)\nif team_filter.strip():\n    data = data[data.team_id ==\
      \ team_filter.strip()]\nif team_name_filter.strip():\n    data = data[data.team_name == team_name_filter.strip()]\n\
      period = data[(data.day >= report_start) & (data.day < report_end)].copy()\nrequests = int(period.requests.sum())\n\
      spend = period.spend_usd.sum(min_count=1)\nknown_latency = period.latency_count.sum()\nknown_ttft\
      \ = period.ttft_count.sum()\ncache_known = period.response_cache_known.sum()\nsummary = pd.DataFrame([dict(\n\
      \    spend_usd=spend, requests=requests, tokens=period.tokens.sum(),\n    cost_per_1k_requests=1000*spend/requests\
      \ if requests else np.nan,\n    failure_rate=period.failures.sum()/(requests-period.unknown_status_requests.sum())\n\
      \        if requests > period.unknown_status_requests.sum() else np.nan,\n    response_cache_hit_rate=period.response_cache_hits.sum()/cache_known\
      \ if cache_known else np.nan,\n    average_latency_seconds=period.latency_seconds_sum.sum(min_count=1)/known_latency\
      \ if known_latency else np.nan,\n    average_ttft_seconds=period.ttft_seconds_sum.sum(min_count=1)/known_ttft\
      \ if known_ttft else np.nan,\n    unknown_cost_requests=period.unknown_cost_requests.sum(),\n  \
      \  unknown_status_requests=period.unknown_status_requests.sum(),\n    observed_days=period.day.nunique())])\n\
      team_spend = period.groupby(['team_id', 'team_name'], as_index=False).agg(\n    spend_usd=('spend_usd',\
      \ lambda s: s.sum(min_count=1)), requests=('requests', 'sum'),\n    tokens=('tokens', 'sum'), failures=('failures',\
      \ 'sum'),\n    observed_days=('day', 'nunique'))\nteam_spend['cost_per_1k_requests'] = 1000*team_spend.spend_usd/team_spend.requests.replace(0,\
      \ np.nan)\nteam_spend['spend_share_pct'] = 100*team_spend.spend_usd/spend if pd.notna(spend) and\
      \ spend > 0 else np.nan\nteam_spend = team_spend.sort_values('spend_usd', ascending=False)\nmodel_spend\
      \ = period.groupby(['team_id', 'model'], as_index=False).agg(\n    spend_usd=('spend_usd', lambda\
      \ s: s.sum(min_count=1)), requests=('requests', 'sum'))\n# Gaps stay unknown. A day without rows\
      \ can mean zero traffic or lost logging.\ndaily = period.groupby(['day', 'team_id'], as_index=False).agg(\n\
      \    spend_usd=('spend_usd', lambda s: s.sum(min_count=1)))\n_grid = pd.MultiIndex.from_product([pd.date_range(report_start,\
      \ report_end-pd.Timedelta(days=1)),\n    sorted(period.team_id.unique())], names=['day', 'team_id'])\n\
      daily = daily.set_index(['day','team_id']).reindex(_grid).reset_index()\n\n# Compare equal 7-day\
      \ windows, excluding today. Require records on all 7 days.\n_recent = data[(data.day >= report_end-pd.Timedelta(days=7))\
      \ & (data.day < report_end)]\n_previous = data[(data.day >= report_end-pd.Timedelta(days=14)) &\
      \ (data.day < report_end-pd.Timedelta(days=7))]\ndef window_totals(frame):\n    return frame.groupby('team_id').agg(spend=('spend_usd',\
      \ lambda s: s.sum(min_count=1)),\n        requests=('requests', 'sum'), days=('day', 'nunique'))\n\
      change = window_totals(_recent).join(window_totals(_previous), lsuffix='_recent', rsuffix='_previous',\
      \ how='outer')\nchange['spend_change_pct'] = 100*(change.spend_recent/change.spend_previous.replace(0,np.nan)-1)\n\
      change['unit_cost_change_pct'] = 100*((change.spend_recent/change.requests_recent.replace(0,np.nan))/\n\
      \    (change.spend_previous/change.requests_previous.replace(0,np.nan))-1)\nchange['review'] = np.where((change.days_recent\
      \ == 7) & (change.days_previous == 7),\n    np.where((change.spend_change_pct > 25) & ((change.spend_recent-change.spend_previous)\
      \ >= 10),\n        'Review cost increase', 'Within rule'), 'Insufficient coverage')\nchange = change.reset_index()\n\
      \n# Keep ROI absent until the business-value input covers the selected period.\nroi = period.groupby('team_id',\
      \ as_index=False).agg(\n    spend_usd=('spend_usd', lambda s: s.sum(min_count=1)),\n    unknown_cost_requests=('unknown_cost_requests',\
      \ 'sum'), observed_days=('day','nunique'))\nroi['roi_pct'] = np.nan\nroi['net_value_usd'] = np.nan\n\
      roi['value_status'] = 'Needs validated business value'\nroi.loc[(roi.unknown_cost_requests > 0)\
      \ | (roi.observed_days < (report_end-report_start).days), 'value_status'] = 'Incomplete spend coverage'\n\
      if team_name_filter.strip():\n    roi['value_status'] = 'ROI needs full team scope'\nif not value_data.empty:\n\
      \    if value_data.team_id.duplicated().any():\n        raise ValueError('Provide one business-value\
      \ row per team and report window.')\n    values = value_data.copy()\n    values[['value_usd','other_cost_usd']]\
      \ = values[['value_usd','other_cost_usd']].apply(pd.to_numeric, errors='raise')\n    if (values[['value_usd','other_cost_usd']]\
      \ < 0).any().any():\n        raise ValueError('Business value and other costs must be nonnegative.')\n\
      \    values['coverage_start'] = pd.to_datetime(values.coverage_start)\n    values['coverage_end']\
      \ = pd.to_datetime(values.coverage_end)\n    for index, row in roi.iterrows():\n        match =\
      \ values[(values.team_id == row.team_id) &\n                       (values.coverage_start == report_start)\
      \ & (values.coverage_end == report_end)]\n        if (len(match) == 1 and pd.notna(row.spend_usd)\
      \ and row.unknown_cost_requests == 0\n                and row.observed_days == (report_end-report_start).days\
      \ and not team_name_filter.strip()):\n            value = match.iloc[0]\n            if pd.notna(value.value_usd)\
      \ and pd.notna(value.other_cost_usd):\n                total_cost = row.spend_usd + value.other_cost_usd\n\
      \                roi.loc[index, 'net_value_usd'] = value.value_usd - total_cost\n              \
      \  roi.loc[index, 'roi_pct'] = 100*(value.value_usd-total_cost)/total_cost if total_cost > 0 else\
      \ np.nan\n                roi.loc[index, 'value_status'] = 'Validated input; excludes untracked\
      \ costs'\n"
- cellId: 3b0d3fac-e8eb-5e3b-a84b-1ec680eca11f
  cellLabel: Spend (USD)
  cellType: METRIC
  config:
    title: Spend (USD)
    valueVariableName: summary
    showComparison: false
    comparisonType: VALUE
    comparisonVariableName: null
    comparisonFormat: null
    comparisonLabel: ''
    displayFormat:
      format: CURRENCY
      currency: USD
      nanFormat: Unknown
      columnType: NUMBER
      showSeparators: true
      numDecimalDigits: 2
      abbreviateLargeNumbers: false
    valueColumn: spend_usd
    valueRowIndex: 0
    valueAggregate: null
    comparisonColumn: null
    comparisonRowIndex: null
    comparisonAggregate: null
    valueResultVariable: null
    comparisonResultVariable: null
    outputResult: false
- cellId: 84a9bca8-98c8-56ee-9312-a4a223311a5b
  cellLabel: Requests
  cellType: METRIC
  config:
    title: Requests
    valueVariableName: summary
    showComparison: false
    comparisonType: VALUE
    comparisonVariableName: null
    comparisonFormat: null
    comparisonLabel: ''
    displayFormat:
      format: NUMBER
      currency: USD
      nanFormat: Unknown
      columnType: NUMBER
      showSeparators: true
      numDecimalDigits: 0
      abbreviateLargeNumbers: false
    valueColumn: requests
    valueRowIndex: 0
    valueAggregate: null
    comparisonColumn: null
    comparisonRowIndex: null
    comparisonAggregate: null
    valueResultVariable: null
    comparisonResultVariable: null
    outputResult: false
- cellId: c12786ae-31ec-5573-9fc6-fdb8c2ba8e35
  cellLabel: Cost per 1K requests
  cellType: METRIC
  config:
    title: Cost per 1K requests
    valueVariableName: summary
    showComparison: false
    comparisonType: VALUE
    comparisonVariableName: null
    comparisonFormat: null
    comparisonLabel: ''
    displayFormat:
      format: CURRENCY
      currency: USD
      nanFormat: Unknown
      columnType: NUMBER
      showSeparators: true
      numDecimalDigits: 2
      abbreviateLargeNumbers: false
    valueColumn: cost_per_1k_requests
    valueRowIndex: 0
    valueAggregate: null
    comparisonColumn: null
    comparisonRowIndex: null
    comparisonAggregate: null
    valueResultVariable: null
    comparisonResultVariable: null
    outputResult: false
- cellId: 4430c1b1-11f5-575b-9e2a-532652592172
  cellLabel: Response cache hit rate
  cellType: METRIC
  config:
    title: Response cache hit rate
    valueVariableName: summary
    showComparison: false
    comparisonType: VALUE
    comparisonVariableName: null
    comparisonFormat: null
    comparisonLabel: ''
    displayFormat:
      format: PERCENT
      currency: USD
      nanFormat: Unknown
      columnType: NUMBER
      showSeparators: true
      numDecimalDigits: 1
      abbreviateLargeNumbers: false
    valueColumn: response_cache_hit_rate
    valueRowIndex: 0
    valueAggregate: null
    comparisonColumn: null
    comparisonRowIndex: null
    comparisonAggregate: null
    valueResultVariable: null
    comparisonResultVariable: null
    outputResult: false
- cellId: d700319a-d863-549c-ba1a-7eba0f047630
  cellLabel: Failure rate (known status)
  cellType: METRIC
  config:
    title: Failure rate (known status)
    valueVariableName: summary
    showComparison: false
    comparisonType: VALUE
    comparisonVariableName: null
    comparisonFormat: null
    comparisonLabel: ''
    displayFormat:
      format: PERCENT
      currency: USD
      nanFormat: Unknown
      columnType: NUMBER
      showSeparators: true
      numDecimalDigits: 2
      abbreviateLargeNumbers: false
    valueColumn: failure_rate
    valueRowIndex: 0
    valueAggregate: null
    comparisonColumn: null
    comparisonRowIndex: null
    comparisonAggregate: null
    valueResultVariable: null
    comparisonResultVariable: null
    outputResult: false
- cellId: a6429d20-b89d-5ac8-a3ab-e1ce488702d1
  cellLabel: Average latency (seconds)
  cellType: METRIC
  config:
    title: Average latency (seconds)
    valueVariableName: summary
    showComparison: false
    comparisonType: VALUE
    comparisonVariableName: null
    comparisonFormat: null
    comparisonLabel: ''
    displayFormat:
      format: NUMBER
      currency: USD
      nanFormat: Unknown
      columnType: NUMBER
      showSeparators: true
      numDecimalDigits: 2
      abbreviateLargeNumbers: false
    valueColumn: average_latency_seconds
    valueRowIndex: 0
    valueAggregate: null
    comparisonColumn: null
    comparisonRowIndex: null
    comparisonAggregate: null
    valueResultVariable: null
    comparisonResultVariable: null
    outputResult: false
- cellId: 31d92c27-a544-5ab7-87f3-0823e03ce3ce
  cellLabel: Average TTFT (seconds)
  cellType: METRIC
  config:
    title: Average TTFT (seconds)
    valueVariableName: summary
    showComparison: false
    comparisonType: VALUE
    comparisonVariableName: null
    comparisonFormat: null
    comparisonLabel: ''
    displayFormat:
      format: NUMBER
      currency: USD
      nanFormat: Unknown
      columnType: NUMBER
      showSeparators: true
      numDecimalDigits: 2
      abbreviateLargeNumbers: false
    valueColumn: average_ttft_seconds
    valueRowIndex: 0
    valueAggregate: null
    comparisonColumn: null
    comparisonRowIndex: null
    comparisonAggregate: null
    valueResultVariable: null
    comparisonResultVariable: null
    outputResult: false
- cellId: 87bd8b11-47f5-590c-bd17-545f0e54195b
  cellLabel: Tokens
  cellType: METRIC
  config:
    title: Tokens
    valueVariableName: summary
    showComparison: false
    comparisonType: VALUE
    comparisonVariableName: null
    comparisonFormat: null
    comparisonLabel: ''
    displayFormat:
      format: NUMBER
      currency: USD
      nanFormat: Unknown
      columnType: NUMBER
      showSeparators: true
      numDecimalDigits: 0
      abbreviateLargeNumbers: true
    valueColumn: tokens
    valueRowIndex: 0
    valueAggregate: null
    comparisonColumn: null
    comparisonRowIndex: null
    comparisonAggregate: null
    valueResultVariable: null
    comparisonResultVariable: null
    outputResult: false
- cellId: 1b51dfdc-87b6-5011-b639-ec174a2eb8e6
  cellLabel: daily_chart
  cellType: CODE
  config:
    source: 'daily_chart = px.line(daily, x=''day'', y=''spend_usd'', color=''team_id'', markers=True,
      title=''Daily team spend (USD)'', template=''plotly_white'')

      daily_chart.update_traces(connectgaps=False)

      daily_chart'
- cellId: 7017e99d-5ac2-5e12-8bdb-abad9e840384
  cellLabel: model_chart
  cellType: CODE
  config:
    source: 'model_chart = px.bar(model_spend, x=''team_id'', y=''spend_usd'', color=''model'', title=''Model
      mix by team (USD)'', template=''plotly_white'')

      model_chart'
- cellId: 8523df68-05e4-52ad-bec2-735200f6f639
  cellLabel: team_table
  cellType: CODE
  config:
    source: "_ranked = team_spend[['team_id', 'team_name', 'spend_usd', 'spend_share_pct',\n         \
      \                'requests', 'cost_per_1k_requests']].rename(columns={\n    'team_id': 'Team ID',\
      \ 'team_name': 'Team name', 'spend_usd': 'Spend (USD)',\n    'spend_share_pct': 'Spend share (%)',\
      \ 'requests': 'Requests',\n    'cost_per_1k_requests': 'USD / 1K requests'})\n_ranked.style.format({'Spend\
      \ (USD)': '${:,.2f}', 'Spend share (%)': '{:.1f}%',\n    'Requests': '{:,.0f}', 'USD / 1K requests':\
      \ '${:.2f}'}, na_rep='Unknown', escape='html').bar(\n    subset=['Spend share (%)'], color='#cdd4ff',\
      \ vmin=0, vmax=100).hide(axis='index').set_table_styles([\n    {'selector': 'th', 'props': [('text-align',\
      \ 'left'), ('border-bottom', '1px solid #d9dee8'), ('padding', '10px')]},\n    {'selector': 'td',\
      \ 'props': [('padding', '10px'), ('border-bottom', '1px solid #edf0f5'), ('font-variant-numeric',\
      \ 'tabular-nums')]}])"
- cellId: 8d61949d-0e4f-50c2-a1d2-9389f8420110
  cellLabel: changes_table
  cellType: CODE
  config:
    source: "_changes = change[['team_id','spend_previous','spend_recent','spend_change_pct',\n      \
      \             'unit_cost_change_pct','review']].rename(columns={\n    'team_id':'Team ID', 'spend_previous':'Previous\
      \ 7 days (USD)',\n    'spend_recent':'Recent 7 days (USD)', 'spend_change_pct':'Spend change (%)',\n\
      \    'unit_cost_change_pct':'Unit cost change (%)', 'review':'Review'})\n_changes.style.format({'Previous\
      \ 7 days (USD)':'${:,.2f}', 'Recent 7 days (USD)':'${:,.2f}',\n    'Spend change (%)':'{:+.1f}%',\
      \ 'Unit cost change (%)':'{:+.1f}%'}, na_rep='Unknown', escape='html').hide(axis='index')"
- cellId: e3f8e8f9-01dd-5dae-80b8-70afa3300118
  cellLabel: roi_table
  cellType: CODE
  config:
    source: roi
- cellId: 338acdeb-6584-5662-a310-6407fc5e3ef3
  cellLabel: quality_table
  cellType: CODE
  config:
    source: summary[['unknown_cost_requests','unknown_status_requests','observed_days']]
- cellId: 101eab1d-0014-51d1-93e7-b698a10c91ff
  cellLabel: Metric notes
  cellType: MARKDOWN
  config:
    source: '## Read the board

      Cache hit rate measures LiteLLM response-cache hits among requests with a known cache flag. Provider
      prompt caching needs provider cache-token data and a pricing baseline. This template does not calculate
      cache savings.


      TTFT covers requests with a completion-start timestamp. Missing values stay unknown. Review data
      coverage before comparing periods. The cost-change rule flags increases above 25% and $10 across
      two complete 7-day windows with records on each day; treat it as a review queue, not an anomaly
      model.


      ROI = (validated business value - LLM spend - other costs) / (LLM spend + other costs). Leave business
      value blank until Finance validates it. Track department ownership and business outcomes in your
      own governed tables. You must verify logging completeness and currency before using this board for
      financial reporting.

      '
appLayout:
  visibleMetadataFields:
  - NAME
  - DESCRIPTION
  - AUTHOR
  - LAST_EDITED
  - LAST_RUN
  - CATEGORIES
  - STATUS
  - TABLE_OF_CONTENTS
  fullWidth: false
  tabs:
  - name: Overview
    rows:
    - columns:
      - start: 0
        end: 120
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 9f2ee2c1-c9b7-5e03-853b-51ba6c25f7dc
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
    - columns:
      - start: 0
        end: 60
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 8c3aac23-e134-5ec1-b0fa-1116428dc185
          sharedFilterId: null
          height: null
          showLabel: true
          explorable: null
      - start: 60
        end: 120
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 777c6579-f9fe-5a12-afb5-ec5dcaae10c6
          sharedFilterId: null
          height: null
          showLabel: true
          explorable: null
    - columns:
      - start: 0
        end: 120
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 8523df68-05e4-52ad-bec2-735200f6f639
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
    - columns:
      - start: 0
        end: 30
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 3b0d3fac-e8eb-5e3b-a84b-1ec680eca11f
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
      - start: 30
        end: 60
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 84a9bca8-98c8-56ee-9312-a4a223311a5b
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
      - start: 60
        end: 90
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: c12786ae-31ec-5573-9fc6-fdb8c2ba8e35
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
      - start: 90
        end: 120
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 4430c1b1-11f5-575b-9e2a-532652592172
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
    - columns:
      - start: 0
        end: 30
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: d700319a-d863-549c-ba1a-7eba0f047630
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
      - start: 30
        end: 60
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: a6429d20-b89d-5ac8-a3ab-e1ce488702d1
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
      - start: 60
        end: 90
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 31d92c27-a544-5ab7-87f3-0823e03ce3ce
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
      - start: 90
        end: 120
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 87bd8b11-47f5-590c-bd17-545f0e54195b
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
    - columns:
      - start: 0
        end: 120
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 1b51dfdc-87b6-5011-b639-ec174a2eb8e6
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
    - columns:
      - start: 0
        end: 120
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 338acdeb-6584-5662-a310-6407fc5e3ef3
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
  - name: Models and cost changes
    rows:
    - columns:
      - start: 0
        end: 120
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 7017e99d-5ac2-5e12-8bdb-abad9e840384
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
    - columns:
      - start: 0
        end: 120
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 8d61949d-0e4f-50c2-a1d2-9389f8420110
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
    - columns:
      - start: 0
        end: 120
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 101eab1d-0014-51d1-93e7-b698a10c91ff
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
  - name: Business ROI
    rows:
    - columns:
      - start: 0
        end: 120
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: e3f8e8f9-01dd-5dae-80b8-70afa3300118
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null
    - columns:
      - start: 0
        end: 120
        elements:
        - showSource: false
          hideOutput: false
          type: CELL
          cellId: 101eab1d-0014-51d1-93e7-b698a10c91ff
          sharedFilterId: null
          height: null
          showLabel: false
          explorable: null