{"templateId":"markdown","sharedDataIds":{"sidebar":"sidebar-@l10n/ja/sidebars.yaml"},"props":{"metadata":{"markdoc":{"tagList":["platform-badges","admonition"]},"type":"markdown"},"seo":{"title":"Treasure AI Studio - データアナリスト向けワークフロー","description":"未知のデータベースの探索からエグゼクティブダッシュボードの構築まで、クエリ・チャート・スキル・ファイルエクスポートを1つのStudioセッションで連携させる完全な分析ワークフローを紹介します。","siteUrl":"https://docs.treasuredata.com","lang":"en-US","llmstxt":{"hide":false,"sections":[{"title":"Table of contents","includeFiles":["**/*"],"excludeFiles":[]}],"excludeFiles":[]}},"dynamicMarkdocComponents":[],"compilationErrors":[],"ast":{"$$mdtype":"Tag","name":"article","attributes":{},"children":[{"$$mdtype":"Tag","name":"Heading","attributes":{"level":1,"id":"データアナリスト向けワークフロー","__idx":0},"children":["データアナリスト向けワークフロー"]},{"$$mdtype":"Tag","name":"PlatformBadges","attributes":{"platforms":["Web","Desktop"]},"children":[]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["このガイドでは、データアナリストがTreasure AI Studio内で実行するワークフローを解説します。曖昧なビジネス上の問いから、洗練されたエクスポート可能なダッシュボードの完成まで、すべてのステップが1つの会話の中で完結します。AIがツール間の移行を担うため、分析そのものに集中できます。"]},{"$$mdtype":"Tag","name":"Admonition","attributes":{"type":"info","name":"注意"},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":["AIはバックグラウンドで複数のツールを連携させています。データ探索にはCLIコマンド、クエリにはTrino SQL、可視化にはHTMLチャート生成、エクスポートにはファイル生成を使用します。すべてのツール呼び出しはチャットストリームに表示され、展開して各ステップを確認できます。AIがどのようにマルチステップのワークフローを計画・実行するかについては、",{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"/ja/products/ai-studio/concepts/orchestrator"},"children":["Agent Orchestrator"]},"をご参照ください。"]}]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":2,"id":"目的","__idx":1},"children":["目的"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["データアナリストが「先四半期のコンバージョン率はどうなったのか？」という問いから、検証済みのステークホルダー向けダッシュボードを1つのセッションで完成できるようにします。通常、分析作業を断片化させているクエリエディタ・可視化ツール・プレゼンテーションソフト間のコンテキストスイッチングを排除します。"]},{"$$mdtype":"Tag","name":"hr","attributes":{},"children":[]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":2,"id":"ある1日の流れ","__idx":2},"children":["ある1日の流れ"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["月曜日の朝、プロダクト担当VPからSlackメッセージが届きます。「コンバージョン率が下がっているようだ。午後2時のリーダーシップミーティングに向けてQ1のパフォーマンス概要をまとめてもらえるか？」"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["通常であれば、クエリエディタを開き、どのテーブルにコンバージョンイベントが格納されているかを思い出し、SQLを書いてデバッグし、結果をスプレッドシートにコピーし、チャートを作成してスライドに貼り付ける、という作業が必要です。2時間のコンテキストスイッチングです。"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["Studioなら、すべてが1つの会話で完結します。目標を伝えると、AIがデータを探索し、最適化されたクエリを書き、チャートを描画し、ダッシュボードを組み立て、資料を生成します。分析の方向性はあなたが決め、ツールの操作はAIが担います。"]},{"$$mdtype":"Tag","name":"hr","attributes":{},"children":[]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":2,"id":"推奨セットアップ","__idx":3},"children":["推奨セットアップ"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["ワークフローを開始する前に、最大限の効率を発揮できるよう環境を設定してください。"]},{"$$mdtype":"Tag","name":"div","attributes":{"className":"md-table-wrapper"},"children":[{"$$mdtype":"Tag","name":"table","attributes":{"className":"md"},"children":[{"$$mdtype":"Tag","name":"thead","attributes":{},"children":[{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"th","attributes":{"data-label":"セットアップ手順"},"children":["セットアップ手順"]},{"$$mdtype":"Tag","name":"th","attributes":{"data-label":"方法"},"children":["方法"]},{"$$mdtype":"Tag","name":"th","attributes":{"data-label":"理由"},"children":["理由"]}]}]},{"$$mdtype":"Tag","name":"tbody","attributes":{},"children":[{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"td","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["プロジェクトを作成する"]}]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["TDコンテキストでデフォルトデータベースを設定した「Analytics」という名前のプロジェクトを作成する"]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["クエリごとにデータベースを指定する手間を省けます"]}]},{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"td","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["スキルを有効化する"]}]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["スキルのドロップダウンで ",{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["trino"]},"、",{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["time-filtering"]},"、",{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["trino-optimizer"]}," を選択する"]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["AIにプラットフォームのSQLパターンを学習させます"]}]},{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"td","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["モデルを選択する"]}]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["まず ",{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["バランス"]},"（Sonnet）を使用し、複雑なJOINには ",{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["プロ"]},"（Opus）に切り替える"]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["速度と推論品質のバランスを取ります"]}]}]}]}]},{"$$mdtype":"Tag","name":"hr","attributes":{},"children":[]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":2,"id":"ステップバイステップのワークフローq1パフォーマンスダッシュボード","__idx":4},"children":["ステップバイステップのワークフロー：Q1パフォーマンスダッシュボード"]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":3,"id":"ステップ1データを探索する","__idx":5},"children":["ステップ1：データを探索する"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["どのテーブルをクエリするかではなく、何を探しているかをAIに伝えることから始めます。"]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"You: I need to analyze Q1 conversion performance. What data do we have\n     related to user signups, purchases, and page views?\n"},"children":[]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["AIがバックグラウンドで行うこと："]}]},{"$$mdtype":"Tag","name":"ol","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"code","attributes":{},"children":["tdx databases"]}," を実行して利用可能なデータベースを一覧表示する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["プロジェクト設定のデフォルトデータベースに対して ",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["tdx tables <database>"]}," を実行する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["有望なテーブルに対して ",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["tdx describe"]}," を実行してスキーマを確認する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["利用可能なデータを要約し、クエリのアプローチを提案する"]}]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["表示される内容："]}]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"AI: [Tool: Bash] tdx databases\n    [Tool: Bash] tdx tables marketing_prod\n    [Tool: Bash] tdx describe marketing_prod.user_events\n    [Tool: Bash] tdx describe marketing_prod.purchases\n\n    I found two relevant tables in marketing_prod:\n\n    | Table | Key Columns | Row Estimate |\n    |-------|-------------|-------------|\n    | user_events | user_id, event_type, time | ~45M rows |\n    | purchases | user_id, amount, product_id, time | ~2.1M rows |\n\n    user_events contains signup, page_view, and add_to_cart events.\n    I can calculate conversion as signups → purchases. Shall I proceed?\n"},"children":[]},{"$$mdtype":"Tag","name":"Admonition","attributes":{"type":"info","name":"プロのヒント"},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":["テーブルが既にわかっている場合は、探索をスキップできます。「Q1のコンバージョンについて ",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["marketing_prod.user_events"]}," と ",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["marketing_prod.purchases"]}," をクエリしてください」と伝えると、AIはデータ探索をスキップしてクエリの構築に直接進みます。"]}]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":3,"id":"ステップ2分析クエリを実行する","__idx":6},"children":["ステップ2：分析クエリを実行する"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["アプローチを確認し、AIにSQLを書かせます。"]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"You: Yes, calculate daily conversion rate (signups to purchases) for Q1.\n     Break it down by week and by traffic source.\n"},"children":[]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["AIが行うこと："]}]},{"$$mdtype":"Tag","name":"ol","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["time-filtering"]}," スキルに従い、",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["td_interval()"]}," を使用してQ1の日付フィルタリングを含むTrinoクエリを構築する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"code","attributes":{},"children":["user_id"]}," をキーに ",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["user_events"]}," と ",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["purchases"]}," をJOINする"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["trino"]}," スキルに従い、パフォーマンスのためにパーティションプルーニングを適用する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"code","attributes":{},"children":["tdx query"]}," 経由でクエリを実行する"]}]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["表示される内容："]}]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"AI: [Tool: Bash] tdx query -e trino \"\n      SELECT\n        TD_TIME_FORMAT(e.time, 'yyyy-ww') AS week,\n        e.source AS traffic_source,\n        COUNT(DISTINCT e.user_id) AS signups,\n        COUNT(DISTINCT p.user_id) AS purchasers,\n        ROUND(COUNT(DISTINCT p.user_id) * 100.0\n              / NULLIF(COUNT(DISTINCT e.user_id), 0), 2) AS conversion_rate\n      FROM marketing_prod.user_events e\n      LEFT JOIN marketing_prod.purchases p\n        ON e.user_id = p.user_id\n        AND TD_INTERVAL(p.time, '-90d')\n      WHERE e.event_type = 'signup'\n        AND TD_INTERVAL(e.time, '-90d')\n      GROUP BY 1, 2\n      ORDER BY 1, 2\n    \"\n\n    Here are the results:\n\n    | Week | Source | Signups | Purchasers | Rate |\n    |------|--------|---------|------------|------|\n    | 2026-01 | organic | 1,240 | 186 | 15.0% |\n    | 2026-01 | paid | 890 | 98 | 11.0% |\n    | ... | ... | ... | ... | ... |\n\n    Conversion rate dropped from 15% in January to 11.2% in March,\n    with paid traffic showing the steepest decline.\n"},"children":[]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":3,"id":"ステップ3結果を可視化する","__idx":7},"children":["ステップ3：結果を可視化する"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["次に、可視化を依頼します。"]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"You: Show me the weekly conversion trend as a line chart,\n     with separate lines for each traffic source.\n"},"children":[]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["AIが行うこと："]}]},{"$$mdtype":"Tag","name":"ol","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":["クエリ結果をチャートのデータ構造（ラベル＋データセット）にフォーマットする"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["チャートを含む自己完結型のHTMLファイルを生成し、",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["open_file"]}," で開く"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["チャートがファイルパネルに表示される"]}]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["表示される内容："]}]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["ファイルパネルに以下を含む折れ線グラフが表示されます。"]},{"$$mdtype":"Tag","name":"ul","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":["X軸：週（2026-01〜2026-13）"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["Y軸：コンバージョン率（%）"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["折れ線：トラフィックソースごとに1本、色分けされている"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["ホバー時に正確な値を表示するツールチップ"]}]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"You: The paid drop is steeper than I expected. Can you add a scatter plot\n     showing spend vs. conversion for paid traffic?\n"},"children":[]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["AIがコストデータをJOINするフォローアップクエリを実行し、散布図を描画します。これでファイルパネルに2つのチャートが表示されます。"]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":3,"id":"ステップ4ダッシュボードを構築する","__idx":8},"children":["ステップ4：ダッシュボードを構築する"]},{"$$mdtype":"Tag","name":"Admonition","attributes":{"type":"warning","name":"近日公開予定"},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":["グリッドダッシュボードは計画中の機能です。このステップで説明するワークフローは、現在のリリースではまだ利用できません。AIに個別のチャートやテーブルを別々のファイルとして生成するよう依頼することは引き続き可能です。"]}]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["分析が完了したら、すべてをグリッドダッシュボードにまとめます。"]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"You: Build a dashboard with:\n     - KPI cards for overall Q1 conversion, total signups, and total revenue\n     - The conversion trend line chart\n     - A bar chart of conversion by source\n     - A table of the top 10 underperforming weeks\n"},"children":[]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["AIが行うこと："]}]},{"$$mdtype":"Tag","name":"ol","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":["KPIの合計値を計算するための追加クエリを実行する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["マルチセルのグリッドレイアウト（上段にKPI、中段にチャート、下段にテーブル）を含むHTMLファイルを生成する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["ファイルパネルでファイルを開く"]}]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["表示される内容："]}]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["以下を含むインタラクティブなダッシュボードが表示されます。"]},{"$$mdtype":"Tag","name":"ul","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":["トレンド矢印付きで合計指標を表示する3つのKPIカード"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["折れ線グラフと棒グラフが横並びで表示される"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["週ごとの詳細を表示するソート可能なテーブルが下部に配置される"]}]},{"$$mdtype":"Tag","name":"Admonition","attributes":{"type":"info","name":"プロのヒント"},"children":[{"$$mdtype":"Tag","name":"p","attributes":{},"children":["ダッシュボードは複数ページに対応する予定です。AIに「地域別の内訳のための2ページ目を追加してください」と依頼すると、ダッシュボードの上部にページセレクターのドロップダウンが表示されます。"]}]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":3,"id":"ステップ5遅いクエリをデバッグする","__idx":9},"children":["ステップ5：遅いクエリをデバッグする"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["分析中にクエリの実行に時間がかかりすぎることがあります。AIが最適化を支援します。"]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"You: That last query took 3 minutes. Can you optimize it?\n"},"children":[]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["AIが行うこと（trino-optimizerスキルに従う）："]}]},{"$$mdtype":"Tag","name":"ol","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":["遅いクエリの構造を分析する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["パーティションプルーニングの欠如やコストの高いJOINを特定する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["最適化パターンを使用してクエリを書き直す（中間結果にはCTAS、大きなカーディナリティには ",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["APPROX_DISTINCT"]},"、複数の ",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["LIKE"]}," 句の代わりに ",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["REGEXP_LIKE"]},"）"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["最適化されたバージョンを実行し、速度向上を報告する"]}]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"AI: The original query scanned all partitions. I've rewritten it with\n    explicit TD_TIME_RANGE bounds and moved the join to a CTE:\n\n    [Tool: Bash] tdx query -e trino \"WITH signups AS (...) ...\"\n\n    Execution time: 12 seconds (was 3 minutes).\n    The main improvement was partition pruning — the original WHERE clause\n    didn't use TD_INTERVAL on the purchases table.\n"},"children":[]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":3,"id":"ステップ6成果物をエクスポートする","__idx":10},"children":["ステップ6：成果物をエクスポートする"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["分析が完了したら、午後2時のミーティング向けの成果物を生成します。"]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"You: Create a PowerPoint deck with these findings for the leadership meeting.\n     Include the charts, key takeaways, and a recommendation slide.\n"},"children":[]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["AIが行うこと："]}]},{"$$mdtype":"Tag","name":"ol","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":["タイトル、エグゼクティブサマリー、コンバージョントレンドチャート、ソース別内訳、パフォーマンスが低い週のテーブル、推奨事項のスライドを含むPPTXを構成する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["作業ディレクトリにファイルを書き込む"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["プレビューボタンとダウンロードボタン付きのファイルカードがチャットに表示される"]}]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"You: Also export the raw data as an Excel file with separate sheets\n     for weekly metrics and source breakdown.\n"},"children":[]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["AIが行うこと："]}]},{"$$mdtype":"Tag","name":"ol","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":["クエリ結果を複数シートのXLSX構造にフォーマットする"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["ファイルを書き込む"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":["ダウンロード可能な別のファイルカードが表示される"]}]},{"$$mdtype":"Tag","name":"hr","attributes":{},"children":[]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":2,"id":"ワークフローのバリエーション","__idx":11},"children":["ワークフローのバリエーション"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["Q1ダッシュボードはワークフローの一例です。Studioで使用できる他の一般的なアナリストワークフローを紹介します。"]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":3,"id":"データ品質監査","__idx":12},"children":["データ品質監査"]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"You: Audit the customer_profiles table for data quality issues —\n     check for nulls, duplicates, and stale records.\n\nAI: 1. [Bash] tdx describe marketing_prod.customer_profiles → スキーマ確認\n    2. [Bash] tdx query → 列ごとのNULL件数\n    3. [Bash] tdx query → user_idの重複検出\n    4. [Bash] tdx query → 鮮度確認（90日以上更新されていないレコード）\n    5. [open_file] 列ごとのNULL率の棒グラフ\n    6. [Text] 優先度付きの改善手順を含むサマリー\n"},"children":[]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":3,"id":"パイプラインのデバッグ","__idx":13},"children":["パイプラインのデバッグ"]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"You: The daily_etl workflow failed last night. What happened?\n\nAI: 1. [Bash] tdx wf sessions daily_etl → 失敗したセッションを特定\n    2. [Bash] tdx wf attempt logs <session_id> → エラーログを読み込む\n    3. [Thinking] 根本原因：テーブルが見つからない（スキーマ変更）\n    4. [Write] 修正を加えた.digファイルを更新\n    5. [Text] 説明と_retryブロックの追加に関する推奨事項\n"},"children":[]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":3,"id":"アドホック探索","__idx":14},"children":["アドホック探索"]},{"$$mdtype":"Tag","name":"CodeBlock","attributes":{"header":{"controls":{"copy":{}}},"source":"You: I just got access to the new clickstream_raw table.\n     What's in it and is it useful for attribution modeling?\n\nAI: 1. [Bash] tdx describe clickstream_raw → スキーマとサンプル確認\n    2. [Bash] tdx query → 行数、日付範囲、イベントタイプ\n    3. [Bash] tdx query → 主要ディメンションのカーディナリティ\n    4. [open_file] イベントタイプの分布\n    5. [Text] 評価：マルチタッチアトリビューションに適している、\n             purchasesテーブルとの推奨JOINストラテジー\n"},"children":[]},{"$$mdtype":"Tag","name":"hr","attributes":{},"children":[]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":2,"id":"aiの透明性オーケストレーターの仕組み","__idx":15},"children":["AIの透明性：オーケストレーターの仕組み"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["このワークフロー全体を通じて、AIはユーザーが確認・操作できる意思決定を行っています。"]},{"$$mdtype":"Tag","name":"div","attributes":{"className":"md-table-wrapper"},"children":[{"$$mdtype":"Tag","name":"table","attributes":{"className":"md"},"children":[{"$$mdtype":"Tag","name":"thead","attributes":{},"children":[{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"th","attributes":{"data-label":"AIの意思決定"},"children":["AIの意思決定"]},{"$$mdtype":"Tag","name":"th","attributes":{"data-label":"ユーザーが確認できる内容"},"children":["ユーザーが確認できる内容"]},{"$$mdtype":"Tag","name":"th","attributes":{"data-label":"上書きする方法"},"children":["上書きする方法"]}]}]},{"$$mdtype":"Tag","name":"tbody","attributes":{},"children":[{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"td","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["クエリ対象のテーブル"]}]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["チャット内の",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["tdx describe"]},"ツール呼び出し"]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["「",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["analytics.events"]},"を代わりに使用してください」"]}]},{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"td","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["SQL構文と関数"]}]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["Bashツールカードに表示される完全なクエリ"]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["「TrinoではなくHiveを使用してください」または「region = 'US'のWHERE句を追加してください」"]}]},{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"td","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["チャートタイプの選択"]}]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["open_file呼び出しで確認できるチャートHTMLファイル"]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["「折れ線グラフではなく棒グラフで表示してください」"]}]},{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"td","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["ダッシュボードのレイアウト"]}]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["Writeツール呼び出しで確認できるYAMLファイル"]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["「KPIを上部の3列行に配置してください」"]}]},{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"td","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["最適化ストラテジー"]}]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["元のクエリと並べて表示される書き直されたクエリ"]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["「元の構造を維持しつつ、パーティションプルーニングを追加してください」"]}]}]}]}]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["AIはプラットフォーム固有のパターンに対してアクティブなスキル（Trino、時間フィルタリング、trino-optimizer）に従いますが、最終的な決定権は常にユーザーにあります。すべてのツール呼び出しは展開可能で、クリックすると正確なコマンドと結果を確認できます。"]},{"$$mdtype":"Tag","name":"hr","attributes":{},"children":[]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":2,"id":"成功基準ビジネス上の成果","__idx":16},"children":["成功基準：ビジネス上の成果"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["このワークフローの最後には、以下が得られます。"]},{"$$mdtype":"Tag","name":"ul","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["検証済みの分析結果"]}," — すべての数値は、チャットに表示された特定のクエリまで追跡可能です"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["インタラクティブなビジュアライゼーション"]}," — ツールチップとレジェンドを備えたチャートで、画面共有中にプレゼンテーションできます"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["ステークホルダー向けダッシュボード"]}," — KPI、チャート、テーブルを含むマルチセルグリッド"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["エクスポート可能な成果物"]}," — ダウンロード可能なPPTXデッキとXLSXデータファイル"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"strong","attributes":{},"children":["監査可能な記録"]}," — 会話全体が分析プロセスのドキュメントとして機能します"]}]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["クエリエディター、BIツール、スライドビルダーを切り替えることなく、すべて1つのセッションで作成できます。"]},{"$$mdtype":"Tag","name":"hr","attributes":{},"children":[]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":2,"id":"確認事項","__idx":17},"children":["確認事項"]},{"$$mdtype":"Tag","name":"p","attributes":{},"children":["このワークフローに従った後、以下のことができるようになります。"]},{"$$mdtype":"Tag","name":"ul","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"input","attributes":{"checked":false,"type":"checkbox","readOnly":true},"children":[]}," 自然言語とAIの",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["tdx describe"]},"呼び出しを使用して、未知のデータベースを探索する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"input","attributes":{"checked":false,"type":"checkbox","readOnly":true},"children":[]}," TD固有の関数を含む最適化されたTrino SQLをリクエストして受け取る"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"input","attributes":{"checked":false,"type":"checkbox","readOnly":true},"children":[]}," クエリ結果からチャートを生成し、そのタイプをカスタマイズする"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"input","attributes":{"checked":false,"type":"checkbox","readOnly":true},"children":[]}," KPI、チャート、テーブルを含むマルチセルグリッドダッシュボードを構築する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"input","attributes":{"checked":false,"type":"checkbox","readOnly":true},"children":[]}," ステークホルダーへの提供のためにPPTXおよびXLSX形式で結果をエクスポートする"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"input","attributes":{"checked":false,"type":"checkbox","readOnly":true},"children":[]}," すべての数値を、それを生成したツール呼び出しまで追跡する"]}]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":2,"id":"トラブルシューティング","__idx":18},"children":["トラブルシューティング"]},{"$$mdtype":"Tag","name":"div","attributes":{"className":"md-table-wrapper"},"children":[{"$$mdtype":"Tag","name":"table","attributes":{"className":"md"},"children":[{"$$mdtype":"Tag","name":"thead","attributes":{},"children":[{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"th","attributes":{"data-label":"問題"},"children":["問題"]},{"$$mdtype":"Tag","name":"th","attributes":{"data-label":"解決策"},"children":["解決策"]}]}]},{"$$mdtype":"Tag","name":"tbody","attributes":{},"children":[{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"td","attributes":{},"children":["AIが存在しないテーブルを参照するSQLを記述する"]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["まずAIに",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["tdx databases"]},"と",{"$$mdtype":"Tag","name":"code","attributes":{},"children":["tdx tables"]},"を実行させて、利用可能なテーブルを確認するよう依頼してください。プロンプトにデータベース名とテーブル名を明示的に指定してください"]}]},{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"td","attributes":{},"children":["クエリ後にチャートがレンダリングされない"]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["クエリがデータを返したか確認してください — ツール呼び出しを展開して確認します。結果セットが空の場合は、クエリフィルターを調整してください"]}]},{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"td","attributes":{},"children":["ダッシュボードのセルが空になっている"]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["グリッドダッシュボードでは、AIが各セルにデータを入力する必要があります。セルが空白の場合は、「クエリの値でKPIセルを埋めてください」とAIに依頼してください"]}]},{"$$mdtype":"Tag","name":"tr","attributes":{},"children":[{"$$mdtype":"Tag","name":"td","attributes":{},"children":["エクスポートが失敗するか、空のファイルが生成される"]},{"$$mdtype":"Tag","name":"td","attributes":{},"children":["現在のセッションでデータが利用可能か確認してください。コンテキストが圧縮されている場合、AIはエクスポート前にクエリを再実行する必要がある場合があります"]}]}]}]}]},{"$$mdtype":"Tag","name":"Heading","attributes":{"level":2,"id":"次のステップ","__idx":19},"children":["次のステップ"]},{"$$mdtype":"Tag","name":"ul","attributes":{},"children":[{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"/ja/products/ai-studio/concepts/orchestrator"},"children":["Agent Orchestrator"]}," — AIの意思決定プロセスを詳しく理解する"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"/ja/products/ai-studio/visualization/charts"},"children":["Charts & Data Visualization"]}," — チャートタイプとダッシュボードの完全なリファレンス"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"/ja/products/ai-studio/query/query-execution"},"children":["Query Execution"]}," — Trinoクエリの仕組みの詳細"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"/ja/products/ai-studio/skills/skills"},"children":["Skills & Marketplace"]}," — SQLおよびオプティマイザースキルの設定"]},{"$$mdtype":"Tag","name":"li","attributes":{},"children":[{"$$mdtype":"Tag","name":"MarkdownLink","attributes":{"href":"/ja/products/ai-studio/files/artifacts"},"children":["File Output"]}," — すべてのファイルエクスポート形式とビューアー機能"]}]}]},"headings":[{"value":"データアナリスト向けワークフロー","id":"データアナリスト向けワークフロー","depth":1},{"value":"目的","id":"目的","depth":2},{"value":"ある1日の流れ","id":"ある1日の流れ","depth":2},{"value":"推奨セットアップ","id":"推奨セットアップ","depth":2},{"value":"ステップバイステップのワークフロー：Q1パフォーマンスダッシュボード","id":"ステップバイステップのワークフローq1パフォーマンスダッシュボード","depth":2},{"value":"ステップ1：データを探索する","id":"ステップ1データを探索する","depth":3},{"value":"ステップ2：分析クエリを実行する","id":"ステップ2分析クエリを実行する","depth":3},{"value":"ステップ3：結果を可視化する","id":"ステップ3結果を可視化する","depth":3},{"value":"ステップ4：ダッシュボードを構築する","id":"ステップ4ダッシュボードを構築する","depth":3},{"value":"ステップ5：遅いクエリをデバッグする","id":"ステップ5遅いクエリをデバッグする","depth":3},{"value":"ステップ6：成果物をエクスポートする","id":"ステップ6成果物をエクスポートする","depth":3},{"value":"ワークフローのバリエーション","id":"ワークフローのバリエーション","depth":2},{"value":"データ品質監査","id":"データ品質監査","depth":3},{"value":"パイプラインのデバッグ","id":"パイプラインのデバッグ","depth":3},{"value":"アドホック探索","id":"アドホック探索","depth":3},{"value":"AIの透明性：オーケストレーターの仕組み","id":"aiの透明性オーケストレーターの仕組み","depth":2},{"value":"成功基準：ビジネス上の成果","id":"成功基準ビジネス上の成果","depth":2},{"value":"確認事項","id":"確認事項","depth":2},{"value":"トラブルシューティング","id":"トラブルシューティング","depth":2},{"value":"次のステップ","id":"次のステップ","depth":2}],"frontmatter":{"seo":{"title":"Treasure AI Studio - データアナリスト向けワークフロー","description":"未知のデータベースの探索からエグゼクティブダッシュボードの構築まで、クエリ・チャート・スキル・ファイルエクスポートを1つのStudioセッションで連携させる完全な分析ワークフローを紹介します。"},"platforms":["Web","Desktop"]},"lastModified":"2026-05-11T06:01:24.000Z","pagePropGetterError":{"message":"","name":""}},"slug":"/ja/products/ai-studio/getting-started/workflow-data-analyst","userData":{"isAuthenticated":false,"teams":["anonymous"]},"isPublic":true}