creation

workflows/Generate weekly supply chain OTIF reports and AI analysis with Notion and GPT-4o-13944/readme-13944.md

@@ -0,0 +1,718 @@
+Generate weekly supply chain OTIF reports and AI analysis with Notion and GPT-4o
+
+https://n8nworkflows.xyz/workflows/generate-weekly-supply-chain-otif-reports-and-ai-analysis-with-notion-and-gpt-4o-13944
+
+
+# Generate weekly supply chain OTIF reports and AI analysis with Notion and GPT-4o
+
+# 1. Workflow Overview
+
+This workflow generates a **weekly OTIF (On-Time In-Full) logistics performance report** from shipment records, enriches the results with **AI-written commentary**, and stores both the KPI data and analysis in **Notion**.
+
+Its main use case is operational or management reporting for retail supply chains, especially where shipment milestone punctuality needs to be tracked week by week and summarised for decision-makers.
+
+## 1.1 Scheduled Input Reception
+
+The workflow starts on a **weekly schedule** and retrieves raw shipment records from a data source represented here by an n8n **Data Table** node.
+
+## 1.2 Weekly KPI Aggregation
+
+It groups all shipment rows by ISO-style week start, then computes weekly KPIs such as:
+
+- total shipments
+- on-time deliveries
+- late shipments
+- average lead time in days and hours
+- on-time rates for transmission, loading, airport, landing, and delivery checkpoints
+
+## 1.3 Weekly AI Analysis
+
+For each aggregated weekly record, an **AI Agent** generates a short business analysis describing performance, the weakest checkpoint, and one recommendation.
+
+## 1.4 Weekly Reporting to Notion
+
+Two Notion outputs are produced for each week:
+
+- a KPI row in the **Daily OTIF Summary** database
+- an AI commentary row in the **AI Analysis** database
+
+## 1.5 Global Cross-Week Summary
+
+After all weekly records are aggregated, a second code node builds a consolidated text prompt containing all weekly indicators. A second AI Agent uses that prompt to generate a **global management summary** across all weeks.
+
+## 1.6 Global Notion Update
+
+The generated overall summary is written back to a dedicated Notion page row labelled **Overall Performance Summary**.
+
+---
+
+# 2. Block-by-Block Analysis
+
+## 2.1 Block: Trigger and Collect Shipment Records
+
+### Overview
+
+This block launches the workflow automatically every week and fetches all shipment records from the configured source. It is the entry point for the entire reporting process.
+
+### Nodes Involved
+
+- Weekly Trigger
+- Collect Shipments from TMS & WMS
+
+### Node Details
+
+#### Weekly Trigger
+
+- **Type and role:** `n8n-nodes-base.scheduleTrigger`; starts the workflow on a weekly cadence.
+- **Configuration choices:** Configured with an interval rule based on `weeks`, meaning the workflow runs once per week.
+- **Key expressions or variables used:** None.
+- **Input and output connections:** No input; outputs to **Collect Shipments from TMS & WMS**.
+- **Version-specific requirements:** Uses type version `1.3`.
+- **Edge cases or failures:**
+  - Workflow may not run if inactive.
+  - Schedule timing depends on instance timezone.
+  - If the workflow execution overlaps with another long-running run, operational contention is possible.
+- **Sub-workflow reference:** None.
+
+#### Collect Shipments from TMS & WMS
+
+- **Type and role:** `n8n-nodes-base.dataTable`; retrieves all shipment rows from an n8n Data Table.
+- **Configuration choices:**
+  - Operation: `get`
+  - Return all rows: enabled
+  - Requires a selected Data Table ID
+- **Key expressions or variables used:** None in the current node configuration.
+- **Input and output connections:** Receives input from **Weekly Trigger**; outputs to **Aggregate by Week**.
+- **Version-specific requirements:** Uses type version `1.1`.
+- **Edge cases or failures:**
+  - Empty table produces no shipment rows downstream.
+  - Wrong or missing table ID prevents retrieval.
+  - If adapted to an external source later, pagination, authentication, or rate limiting may become relevant.
+- **Sub-workflow reference:** None.
+
+---
+
+## 2.2 Block: Aggregate Shipments and Compute KPIs
+
+### Overview
+
+This block transforms raw shipment records into one record per week. It calculates all operational KPIs used later by the AI analysis and Notion reporting branches.
+
+### Nodes Involved
+
+- Aggregate by Week
+
+### Node Details
+
+#### Aggregate by Week
+
+- **Type and role:** `n8n-nodes-base.code`; custom JavaScript aggregation logic.
+- **Configuration choices:**
+  - Reads all incoming shipment items with `$input.all()`.
+  - Groups records by Monday-based week key derived from `Order_Time`.
+  - Calculates:
+    - `Name` as `Week <number>`
+    - `weekStart`, `weekEnd`
+    - `totalShipments`
+    - `onTimeDeliveries`
+    - `avgLeadTimeDays`
+    - `avgLeadTimeHrs`
+    - checkpoint percentages:
+      - `transmissionOnTime`
+      - `loadingOnTime`
+      - `airportOnTime`
+      - `landingOnTime`
+      - `deliveryOnTime`
+    - `lateShipments`
+  - Sorts results ascending by `weekStart`.
+- **Key expressions or variables used:**
+  - `s.Order_Time`
+  - `r.Delivery_OnTime`
+  - `r.LT_Days`
+  - `r.LT_Hours`
+  - `r.Transmission_OnTime`
+  - `r.Loading_OnTime`
+  - `r.Airport_OnTime`
+  - `r.Landing_OnTime`
+- **Input and output connections:** Receives input from **Collect Shipments from TMS & WMS**; outputs to:
+  - **AI Agent Weekly Performance Summary**
+  - **Prepare Global Summary Prompt with Indicators**
+- **Version-specific requirements:** Uses code node type version `2`.
+- **Edge cases or failures:**
+  - If no shipments are received, the node returns a single item with `{ error: 'No shipments found', count: 0 }`. Downstream nodes are not designed explicitly for that shape, so later expressions may fail or produce poor summaries.
+  - Invalid or missing `Order_Time` can create invalid date values.
+  - Missing numeric fields are partly tolerated with `(value || 0)` logic for lead times, but boolean metrics assume proper `true` values.
+  - Division-by-zero is avoided in practice because KPI calculation only occurs for populated weekly groups, but malformed records could still create inconsistent outputs.
+  - Week number logic is custom and may differ slightly from strict ISO week expectations in edge calendar cases.
+- **Sub-workflow reference:** None.
+
+---
+
+## 2.3 Block: AI Agent Generates a Per-Week Performance Analysis
+
+### Overview
+
+This block sends each weekly KPI object to an AI Agent, which produces a concise operational assessment. A dedicated OpenAI chat model is attached as the language model backend.
+
+### Nodes Involved
+
+- OpenAI Chat Model
+- AI Agent Weekly Performance Summary
+
+### Node Details
+
+#### OpenAI Chat Model
+
+- **Type and role:** `@n8n/n8n-nodes-langchain.lmChatOpenAi`; language model provider for the weekly AI agent.
+- **Configuration choices:**
+  - Model: `gpt-4o-mini`
+  - Max tokens: `300`
+  - Temperature: `0.3`
+- **Key expressions or variables used:** None directly in this node.
+- **Input and output connections:** Connected via `ai_languageModel` to **AI Agent Weekly Performance Summary**.
+- **Version-specific requirements:** Uses type version `1.2`.
+- **Edge cases or failures:**
+  - Missing or invalid OpenAI credentials.
+  - Model access restrictions on the account.
+  - Token or rate-limit issues.
+  - Temporary API failures.
+- **Sub-workflow reference:** None.
+
+#### AI Agent Weekly Performance Summary
+
+- **Type and role:** `@n8n/n8n-nodes-langchain.agent`; generates a textual weekly logistics analysis.
+- **Configuration choices:**
+  - Prompt type: defined directly in the node.
+  - System message instructs the agent to act as a supply chain OTIF analyst, be concise and actionable, and use British spelling.
+  - The main prompt injects KPI fields from the current weekly JSON item and requests:
+    - overall performance assessment
+    - weakest checkpoint
+    - one specific recommendation
+    - 3–4 concise sentences
+- **Key expressions or variables used:**
+  - `{{ $json.Name }}`
+  - `{{ $json.weekStart }}`
+  - `{{ $json.weekEnd }}`
+  - `{{ $json.totalShipments }}`
+  - `{{ $json.onTimeDeliveries }}`
+  - `{{ $json.lateShipments }}`
+  - `{{ Math.round($json.onTimeDeliveries / $json.totalShipments * 1000) / 10 }}`
+  - `{{ $json.avgLeadTimeDays }}`
+  - `{{ $json.avgLeadTimeHrs }}`
+  - checkpoint percentages from `$json`
+- **Input and output connections:**
+  - Main input from **Aggregate by Week**
+  - AI language model input from **OpenAI Chat Model**
+  - Main outputs to:
+    - **Fill the report**
+    - **Create Weekly Performance Card**
+- **Version-specific requirements:** Uses type version `1.7`.
+- **Edge cases or failures:**
+  - If upstream data is missing expected KPI fields, the prompt may render incorrectly.
+  - If `totalShipments` is zero or undefined, the OTIF percentage expression can become invalid.
+  - LLM output is non-deterministic, though reduced by low temperature.
+  - Output format depends on the agent’s standard `output` field; downstream nodes assume this field exists.
+- **Sub-workflow reference:** None.
+
+---
+
+## 2.4 Block: Push Weekly KPIs and AI Analysis Cards to Notion
+
+### Overview
+
+This block stores the weekly results in Notion. One node writes KPI metrics to a scorecard database, while another writes the AI-generated weekly commentary to a separate analysis database.
+
+### Nodes Involved
+
+- Fill the report
+- Create Weekly Performance Card
+
+### Node Details
+
+#### Fill the report
+
+- **Type and role:** `n8n-nodes-base.notion`; creates a page in a Notion database for weekly KPI reporting.
+- **Configuration choices:**
+  - Resource: `databasePage`
+  - Operation: create
+  - Title and `Name|title` come from `Aggregate by Week`
+  - Icon: truck emoji `🚚`
+  - Writes numeric fields for all calculated KPI metrics
+  - Writes a date range using `weekStart` and `weekEnd`
+  - Includes an `AI Analysis|rich_text` property, but it is left empty in the current configuration
+  - Database target is the Notion database labelled `Daily OTIF Summary`
+- **Key expressions or variables used:**
+  - `{{ $('Aggregate by Week').item.json.Name }}`
+  - `{{ $('Aggregate by Week').item.json.airportOnTime }}`
+  - `{{ $('Aggregate by Week').item.json.avgLeadTimeDays }}`
+  - `{{ $('Aggregate by Week').item.json.avgLeadTimeHrs }}`
+  - `{{ $('Aggregate by Week').item.json.weekEnd }}`
+  - `{{ $('Aggregate by Week').item.json.weekStart }}`
+  - `{{ $('Aggregate by Week').item.json.deliveryOnTime }}`
+  - `{{ $('Aggregate by Week').item.json.landingOnTime }}`
+  - `{{ $('Aggregate by Week').item.json.lateShipments }}`
+  - `{{ $('Aggregate by Week').item.json.loadingOnTime }}`
+  - `{{ $('Aggregate by Week').item.json.onTimeDeliveries }}`
+  - `{{ $('Aggregate by Week').item.json.totalShipments }}`
+  - `{{ $('Aggregate by Week').item.json.transmissionOnTime }}`
+- **Input and output connections:** Receives input from **AI Agent Weekly Performance Summary**; no downstream connections.
+- **Version-specific requirements:** Uses Notion node type version `2.2`.
+- **Edge cases or failures:**
+  - Missing Notion credentials or insufficient database permissions.
+  - Database schema mismatch: every configured property name must exist exactly in Notion.
+  - Cross-node `.item` references can fail if item linking changes unexpectedly.
+  - The AI Analysis property is present but not populated, which may be intentional or an incomplete design.
+- **Sub-workflow reference:** None.
+
+#### Create Weekly Performance Card
+
+- **Type and role:** `n8n-nodes-base.notion`; creates a page in a second Notion database for AI commentary.
+- **Configuration choices:**
+  - Resource: `databasePage`
+  - Operation: create
+  - Icon: chart emoji `📊`
+  - Database target labelled `AI Analysis`
+  - Writes:
+    - `Name|title` from aggregated week name
+    - `Weekly Analysis|rich_text` from AI output
+    - `Type|select` as `Weekly Analysis`
+    - `Updated|date` with current execution date generated by Notion node behavior for date property when included without explicit value
+- **Key expressions or variables used:**
+  - `{{ $('Aggregate by Week').item.json.Name }}`
+  - `{{ $json.output }}`
+- **Input and output connections:** Receives input from **AI Agent Weekly Performance Summary**; no downstream connections.
+- **Version-specific requirements:** Uses Notion node type version `2.2`.
+- **Edge cases or failures:**
+  - Same Notion auth and schema concerns as above.
+  - If the AI agent output field is absent, the rich text field will be blank or error.
+  - The `Type` select option must already exist or be supported by the Notion API configuration.
+- **Sub-workflow reference:** None.
+
+---
+
+## 2.5 Block: AI Agent Generates a Global Cross-Week Performance Summary
+
+### Overview
+
+This block consolidates all weekly KPI rows into one textual management prompt, then asks a second AI Agent to create a cross-week summary with trends, bottlenecks, best/worst weeks, and recommendations.
+
+### Nodes Involved
+
+- Prepare Global Summary Prompt with Indicators
+- OpenAI Chat Model Global
+- AI Agent Global Performance Summary
+
+### Node Details
+
+#### Prepare Global Summary Prompt with Indicators
+
+- **Type and role:** `n8n-nodes-base.code`; builds a single prompt string covering every aggregated week.
+- **Configuration choices:**
+  - Reads all weekly items using `$input.all()`
+  - Sorts by `weekStart`
+  - Creates a `globalPrompt` text block containing weekly shipment counts, OTIF, lead times, and checkpoint rates
+  - Returns a single item with `{ globalPrompt: summary }`
+- **Key expressions or variables used:**
+  - `w.weekStart`
+  - `w.weekEnd`
+  - `w.onTimeDeliveries`
+  - `w.totalShipments`
+  - `w.lateShipments`
+  - `w.avgLeadTimeDays`
+  - `w.transmissionOnTime`
+  - `w.loadingOnTime`
+  - `w.airportOnTime`
+  - `w.landingOnTime`
+  - `w.deliveryOnTime`
+- **Input and output connections:** Receives input from **Aggregate by Week**; outputs to **AI Agent Global Performance Summary**.
+- **Version-specific requirements:** Uses code node type version `2`.
+- **Edge cases or failures:**
+  - If upstream contains an error item instead of KPI items, the generated prompt will be incomplete or malformed.
+  - Division-based OTIF calculation assumes `totalShipments > 0`.
+  - Very long historical data could make the prompt large; token growth may matter if many weeks are included.
+- **Sub-workflow reference:** None.
+
+#### OpenAI Chat Model Global
+
+- **Type and role:** `@n8n/n8n-nodes-langchain.lmChatOpenAi`; language model provider for the global AI agent.
+- **Configuration choices:**
+  - Model: `gpt-4o-mini`
+  - Max tokens: `600`
+  - Temperature: `0.3`
+- **Key expressions or variables used:** None directly.
+- **Input and output connections:** Connected via `ai_languageModel` to **AI Agent Global Performance Summary**.
+- **Version-specific requirements:** Uses type version `1.2`.
+- **Edge cases or failures:**
+  - Same OpenAI credential, access, and rate-limit risks as the weekly model node.
+  - If prompts become too long, truncation or quality degradation may occur.
+- **Sub-workflow reference:** None.
+
+#### AI Agent Global Performance Summary
+
+- **Type and role:** `@n8n/n8n-nodes-langchain.agent`; produces a multi-week management summary.
+- **Configuration choices:**
+  - Prompt type: defined directly in the node.
+  - Prompt begins with `{{ $json.globalPrompt }}`
+  - Explicitly asks for:
+    1. overall trend
+    2. most consistent bottleneck checkpoint
+    3. best and worst weeks with reasons
+    4. two recommendations
+    5. outlook
+  - Output format requested as concise bullet points without markdown headers
+  - System message sets role, scope, and British spelling
+- **Key expressions or variables used:**
+  - `{{ $json.globalPrompt }}`
+- **Input and output connections:**
+  - Main input from **Prepare Global Summary Prompt with Indicators**
+  - AI language model input from **OpenAI Chat Model Global**
+  - Main output to **Update Global Performance Summary**
+- **Version-specific requirements:** Uses type version `1.7`.
+- **Edge cases or failures:**
+  - If the summary prompt is empty, the agent output will be low quality.
+  - Output shape must contain `output` for downstream Notion update.
+  - Large weekly histories can increase latency and cost.
+- **Sub-workflow reference:** None.
+
+---
+
+## 2.6 Block: Update the Global Performance Summary Card in Notion
+
+### Overview
+
+This block writes the final cross-week AI analysis into a dedicated Notion page. It acts as the management-facing rollup of the whole workflow.
+
+### Nodes Involved
+
+- Update Global Performance Summary
+
+### Node Details
+
+#### Update Global Performance Summary
+
+- **Type and role:** `n8n-nodes-base.notion`; updates an existing Notion database page rather than creating a new one.
+- **Configuration choices:**
+  - Resource: `databasePage`
+  - Operation: `update`
+  - Target page selected by explicit `pageId`
+  - Sets:
+    - `Name|title` = `Overall Performance Summary`
+    - `Weekly Analysis|rich_text` = AI output
+    - `Type|select` = `Global Summary`
+    - `Updated|date`
+- **Key expressions or variables used:**
+  - `{{ $json.output }}`
+- **Input and output connections:** Receives input from **AI Agent Global Performance Summary**; no downstream node.
+- **Version-specific requirements:** Uses Notion node type version `2.2`.
+- **Edge cases or failures:**
+  - `pageId` must point to an existing page in the correct database.
+  - Notion permissions must allow update access.
+  - Property names and types must match the database schema exactly.
+  - If the page is deleted or moved, updates will fail.
+- **Sub-workflow reference:** None.
+
+---
+
+## 2.7 Non-Executable Documentation Nodes
+
+### Overview
+
+These nodes are visual annotations only. They explain setup, process stages, and provide a video link.
+
+### Nodes Involved
+
+- Sticky Note
+- Sticky Note1
+- Sticky Note2
+- Sticky Note3
+- Sticky Note4
+- Sticky Note5
+- Sticky Note6
+- Sticky Note7
+
+### Node Details
+
+All sticky notes are of type `n8n-nodes-base.stickyNote` and have no execution role, inputs, or outputs.
+
+#### Sticky Note
+
+- General overview of the whole workflow.
+- Includes setup checklist:
+  - connect data source
+  - add OpenAI credentials
+  - add Notion credentials
+  - verify Notion database IDs
+- Includes customisation notes:
+  - adjust AI prompts
+  - modify KPI logic in aggregation code
+
+#### Sticky Note1
+
+- Section label: `1. Trigger and collect shipment records`
+
+#### Sticky Note2
+
+- Section label: `2. Aggregate shipments and compute KPIs`
+
+#### Sticky Note3
+
+- Section label: `3. AI Agent generates a per-week performance analysis`
+
+#### Sticky Note4
+
+- Section label: `4. Push weekly KPIs and AI analysis cards to Notion`
+
+#### Sticky Note5
+
+- Section label: `5. AI Agent generates a global cross-week performance summary`
+
+#### Sticky Note6
+
+- Section label: `6. Update the global performance summary card in Notion`
+
+#### Sticky Note7
+
+- Contains a video link:
+  - `https://www.youtube.com/watch?v=tOT8XhQ7eB8`
+
+---
+
+# 3. Summary Table
+
+| Node Name | Node Type | Functional Role | Input Node(s) | Output Node(s) | Sticky Note |
+|---|---|---|---|---|---|
+| Weekly Trigger | n8n-nodes-base.scheduleTrigger | Weekly workflow entry point |  | Collect Shipments from TMS & WMS | ## Supply Chain OTIF Performance Scorecard for Retail Logistics<br>Automatically aggregate shipment data, compute weekly OTIF KPIs, generate AI-powered performance analysis, and push everything to a Notion dashboard.<br>### How it Works<br>1. **Collect** raw shipment records from your TMS/WMS data source.<br>2. **Aggregate** shipments by ISO week and compute KPIs (OTIF rate, lead times, checkpoint on-time rates).<br>3. **AI Weekly Analysis** generates a per-week performance comment using an AI Agent.<br>4. **Push to Notion** creates one row per week in the OTIF Summary database and one card per week in the AI Analysis database.<br>5. **AI Global Analysis** generates a cross-week summary with trends and recommendations.<br>6. **Update Global Card** pushes the overall summary to a dedicated Notion database row.<br>### Setup<br>- [ ] Connect your **data source** (DataTable, database, or API) to the "Collect Shipments" node<br>- [ ] Add your **OpenAI API Key** to both OpenAI Chat Model nodes<br>- [ ] Add your **Notion API credentials** to all Notion nodes<br>- [ ] Verify the **Notion database IDs** match your workspace<br>### Customisation<br>- Adjust the AI prompts to change the analysis style or language<br>- Modify the Aggregate by Week code to add custom KPI<br>## 1. Trigger and collect shipment records |
+| Collect Shipments from TMS & WMS | n8n-nodes-base.dataTable | Fetch all shipment records | Weekly Trigger | Aggregate by Week | ## Supply Chain OTIF Performance Scorecard for Retail Logistics<br>Automatically aggregate shipment data, compute weekly OTIF KPIs, generate AI-powered performance analysis, and push everything to a Notion dashboard.<br>### How it Works<br>1. **Collect** raw shipment records from your TMS/WMS data source.<br>2. **Aggregate** shipments by ISO week and compute KPIs (OTIF rate, lead times, checkpoint on-time rates).<br>3. **AI Weekly Analysis** generates a per-week performance comment using an AI Agent.<br>4. **Push to Notion** creates one row per week in the OTIF Summary database and one card per week in the AI Analysis database.<br>5. **AI Global Analysis** generates a cross-week summary with trends and recommendations.<br>6. **Update Global Card** pushes the overall summary to a dedicated Notion database row.<br>### Setup<br>- [ ] Connect your **data source** (DataTable, database, or API) to the "Collect Shipments" node<br>- [ ] Add your **OpenAI API Key** to both OpenAI Chat Model nodes<br>- [ ] Add your **Notion API credentials** to all Notion nodes<br>- [ ] Verify the **Notion database IDs** match your workspace<br>### Customisation<br>- Adjust the AI prompts to change the analysis style or language<br>- Modify the Aggregate by Week code to add custom KPI<br>## 1. Trigger and collect shipment records |
+| Aggregate by Week | n8n-nodes-base.code | Group shipments by week and compute KPIs | Collect Shipments from TMS & WMS | AI Agent Weekly Performance Summary; Prepare Global Summary Prompt with Indicators | ## Supply Chain OTIF Performance Scorecard for Retail Logistics<br>Automatically aggregate shipment data, compute weekly OTIF KPIs, generate AI-powered performance analysis, and push everything to a Notion dashboard.<br>### How it Works<br>1. **Collect** raw shipment records from your TMS/WMS data source.<br>2. **Aggregate** shipments by ISO week and compute KPIs (OTIF rate, lead times, checkpoint on-time rates).<br>3. **AI Weekly Analysis** generates a per-week performance comment using an AI Agent.<br>4. **Push to Notion** creates one row per week in the OTIF Summary database and one card per week in the AI Analysis database.<br>5. **AI Global Analysis** generates a cross-week summary with trends and recommendations.<br>6. **Update Global Card** pushes the overall summary to a dedicated Notion database row.<br>### Setup<br>- [ ] Connect your **data source** (DataTable, database, or API) to the "Collect Shipments" node<br>- [ ] Add your **OpenAI API Key** to both OpenAI Chat Model nodes<br>- [ ] Add your **Notion API credentials** to all Notion nodes<br>- [ ] Verify the **Notion database IDs** match your workspace<br>### Customisation<br>- Adjust the AI prompts to change the analysis style or language<br>- Modify the Aggregate by Week code to add custom KPI<br>## 2. Aggregate shipments and compute KPIs |
+| OpenAI Chat Model | @n8n/n8n-nodes-langchain.lmChatOpenAi | LLM backend for weekly AI analysis |  | AI Agent Weekly Performance Summary | ## 3. AI Agent generates a per-week performance analysis |
+| AI Agent Weekly Performance Summary | @n8n/n8n-nodes-langchain.agent | Generate weekly AI commentary | Aggregate by Week; OpenAI Chat Model | Fill the report; Create Weekly Performance Card | ## Supply Chain OTIF Performance Scorecard for Retail Logistics<br>Automatically aggregate shipment data, compute weekly OTIF KPIs, generate AI-powered performance analysis, and push everything to a Notion dashboard.<br>### How it Works<br>1. **Collect** raw shipment records from your TMS/WMS data source.<br>2. **Aggregate** shipments by ISO week and compute KPIs (OTIF rate, lead times, checkpoint on-time rates).<br>3. **AI Weekly Analysis** generates a per-week performance comment using an AI Agent.<br>4. **Push to Notion** creates one row per week in the OTIF Summary database and one card per week in the AI Analysis database.<br>5. **AI Global Analysis** generates a cross-week summary with trends and recommendations.<br>6. **Update Global Card** pushes the overall summary to a dedicated Notion database row.<br>### Setup<br>- [ ] Connect your **data source** (DataTable, database, or API) to the "Collect Shipments" node<br>- [ ] Add your **OpenAI API Key** to both OpenAI Chat Model nodes<br>- [ ] Add your **Notion API credentials** to all Notion nodes<br>- [ ] Verify the **Notion database IDs** match your workspace<br>### Customisation<br>- Adjust the AI prompts to change the analysis style or language<br>- Modify the Aggregate by Week code to add custom KPI<br>## 3. AI Agent generates a per-week performance analysis |
+| Fill the report | n8n-nodes-base.notion | Create weekly KPI page in Notion OTIF database | AI Agent Weekly Performance Summary |  | ## 4. Push weekly KPIs and AI analysis cards to Notion |
+| Create Weekly Performance Card | n8n-nodes-base.notion | Create weekly AI analysis page in Notion analysis database | AI Agent Weekly Performance Summary |  | ## 4. Push weekly KPIs and AI analysis cards to Notion |
+| Prepare Global Summary Prompt with Indicators | n8n-nodes-base.code | Build a consolidated multi-week prompt | Aggregate by Week | AI Agent Global Performance Summary | ## Supply Chain OTIF Performance Scorecard for Retail Logistics<br>Automatically aggregate shipment data, compute weekly OTIF KPIs, generate AI-powered performance analysis, and push everything to a Notion dashboard.<br>### How it Works<br>1. **Collect** raw shipment records from your TMS/WMS data source.<br>2. **Aggregate** shipments by ISO week and compute KPIs (OTIF rate, lead times, checkpoint on-time rates).<br>3. **AI Weekly Analysis** generates a per-week performance comment using an AI Agent.<br>4. **Push to Notion** creates one row per week in the OTIF Summary database and one card per week in the AI Analysis database.<br>5. **AI Global Analysis** generates a cross-week summary with trends and recommendations.<br>6. **Update Global Card** pushes the overall summary to a dedicated Notion database row.<br>### Setup<br>- [ ] Connect your **data source** (DataTable, database, or API) to the "Collect Shipments" node<br>- [ ] Add your **OpenAI API Key** to both OpenAI Chat Model nodes<br>- [ ] Add your **Notion API credentials** to all Notion nodes<br>- [ ] Verify the **Notion database IDs** match your workspace<br>### Customisation<br>- Adjust the AI prompts to change the analysis style or language<br>- Modify the Aggregate by Week code to add custom KPI<br>## 5. AI Agent generates a global cross-week performance summary |
+| OpenAI Chat Model Global | @n8n/n8n-nodes-langchain.lmChatOpenAi | LLM backend for global AI summary |  | AI Agent Global Performance Summary | ## 5. AI Agent generates a global cross-week performance summary |
+| AI Agent Global Performance Summary | @n8n/n8n-nodes-langchain.agent | Generate cross-week management summary | Prepare Global Summary Prompt with Indicators; OpenAI Chat Model Global | Update Global Performance Summary | ## Supply Chain OTIF Performance Scorecard for Retail Logistics<br>Automatically aggregate shipment data, compute weekly OTIF KPIs, generate AI-powered performance analysis, and push everything to a Notion dashboard.<br>### How it Works<br>1. **Collect** raw shipment records from your TMS/WMS data source.<br>2. **Aggregate** shipments by ISO week and compute KPIs (OTIF rate, lead times, checkpoint on-time rates).<br>3. **AI Weekly Analysis** generates a per-week performance comment using an AI Agent.<br>4. **Push to Notion** creates one row per week in the OTIF Summary database and one card per week in the AI Analysis database.<br>5. **AI Global Analysis** generates a cross-week summary with trends and recommendations.<br>6. **Update Global Card** pushes the overall summary to a dedicated Notion database row.<br>### Setup<br>- [ ] Connect your **data source** (DataTable, database, or API) to the "Collect Shipments" node<br>- [ ] Add your **OpenAI API Key** to both OpenAI Chat Model nodes<br>- [ ] Add your **Notion API credentials** to all Notion nodes<br>- [ ] Verify the **Notion database IDs** match your workspace<br>### Customisation<br>- Adjust the AI prompts to change the analysis style or language<br>- Modify the Aggregate by Week code to add custom KPI<br>## 5. AI Agent generates a global cross-week performance summary |
+| Update Global Performance Summary | n8n-nodes-base.notion | Update a dedicated Notion page with overall AI summary | AI Agent Global Performance Summary |  | ## 6. Update the global performance summary card in Notion<br>## [Tutorial](https://www.youtube.com/watch?v=tOT8XhQ7eB8)<br>@[youtube](tOT8XhQ7eB8) |
+| Sticky Note | n8n-nodes-base.stickyNote | Visual documentation |  |  |  |
+| Sticky Note1 | n8n-nodes-base.stickyNote | Visual section label |  |  |  |
+| Sticky Note2 | n8n-nodes-base.stickyNote | Visual section label |  |  |  |
+| Sticky Note3 | n8n-nodes-base.stickyNote | Visual section label |  |  |  |
+| Sticky Note4 | n8n-nodes-base.stickyNote | Visual section label |  |  |  |
+| Sticky Note5 | n8n-nodes-base.stickyNote | Visual section label |  |  |  |
+| Sticky Note6 | n8n-nodes-base.stickyNote | Visual section label |  |  |  |
+| Sticky Note7 | n8n-nodes-base.stickyNote | Visual external resource note |  |  |  |
+
+---
+
+# 4. Reproducing the Workflow from Scratch
+
+1. **Create a new workflow** in n8n.
+
+2. **Add a Schedule Trigger node**.
+   - Node type: `Schedule Trigger`
+   - Configure it to run every week.
+   - Keep timezone considerations in mind based on your n8n instance settings.
+   - Rename it to **Weekly Trigger**.
+
+3. **Add a shipment source node**.
+   - In this workflow, use a `Data Table` node.
+   - Rename it to **Collect Shipments from TMS & WMS**.
+   - Set:
+     - Operation: `Get`
+     - Return All: enabled
+     - Data Table ID: select your shipment table
+   - Connect **Weekly Trigger → Collect Shipments from TMS & WMS**.
+
+4. **Ensure your source data contains these fields** for each shipment row:
+   - `Order_Time`
+   - `Delivery_OnTime`
+   - `LT_Days`
+   - `LT_Hours`
+   - `Transmission_OnTime`
+   - `Loading_OnTime`
+   - `Airport_OnTime`
+   - `Landing_OnTime`
+   - Optional but useful: any shipment identifiers for future extensions
+
+5. **Add a Code node** for weekly aggregation.
+   - Rename it to **Aggregate by Week**.
+   - Paste the aggregation logic that:
+     - loads all shipment rows
+     - groups by Monday-based week
+     - computes KPI metrics
+     - returns one item per week
+   - Connect **Collect Shipments from TMS & WMS → Aggregate by Week**.
+
+6. **Add an OpenAI Chat Model node** for weekly analysis.
+   - Node type: `OpenAI Chat Model` from LangChain integration
+   - Rename it to **OpenAI Chat Model**
+   - Configure:
+     - Model: `gpt-4o-mini`
+     - Max Tokens: `300`
+     - Temperature: `0.3`
+   - Attach valid **OpenAI API credentials**.
+
+7. **Add an AI Agent node** for weekly analysis.
+   - Rename it to **AI Agent Weekly Performance Summary**.
+   - Set prompt mode to define text manually.
+   - Add this logic in plain form:
+     - include week name, week start/end, shipment counts, lead time, and checkpoint percentages
+     - ask for a brief 3–4 sentence analysis
+     - request overall assessment, weakest checkpoint, and one recommendation
+   - Add a system message instructing the model to act as a supply chain OTIF analyst, be concise, actionable, and use British spelling.
+   - Connect:
+     - **Aggregate by Week → AI Agent Weekly Performance Summary** via main connection
+     - **OpenAI Chat Model → AI Agent Weekly Performance Summary** via AI language model connection
+
+8. **Add a Notion node** for weekly KPI storage.
+   - Rename it to **Fill the report**.
+   - Resource: `Database Page`
+   - Operation: create
+   - Select your Notion credentials.
+   - Select the target database, equivalent to **Daily OTIF Summary**.
+   - Configure page icon as emoji `🚚`.
+   - Map these properties:
+     - `Name` as title from the aggregated week name
+     - `Airport On-Time %`
+     - `Avg Lead Time (days)`
+     - `Avg Lead Time (hrs)`
+     - `Date` as a date range using week start and week end
+     - `Delivery On-Time %`
+     - `Landing On-Time %`
+     - `Late Shipments`
+     - `Loading On-Time %`
+     - `On-Time Deliveries`
+     - `Total Shipments`
+     - `Transmission On-Time %`
+     - `AI Analysis` rich text property if desired
+   - Use expressions referencing **Aggregate by Week** values, not the AI output.
+   - Connect **AI Agent Weekly Performance Summary → Fill the report**.
+
+9. **Create the first Notion database schema** before testing.
+   - Database name can be `Daily OTIF Summary`.
+   - Required properties and types:
+     - `Name` → Title
+     - `Airport On-Time %` → Number
+     - `Avg Lead Time (days)` → Number
+     - `Avg Lead Time (hrs)` → Number
+     - `Date` → Date
+     - `Delivery On-Time %` → Number
+     - `Landing On-Time %` → Number
+     - `Late Shipments` → Number
+     - `Loading On-Time %` → Number
+     - `On-Time Deliveries` → Number
+     - `Total Shipments` → Number
+     - `Transmission On-Time %` → Number
+     - `AI Analysis` → Rich text
+
+10. **Add a second Notion node** for weekly AI commentary.
+    - Rename it to **Create Weekly Performance Card**.
+    - Resource: `Database Page`
+    - Operation: create
+    - Choose the second Notion database, equivalent to **AI Analysis**.
+    - Set icon as emoji `📊`.
+    - Map:
+      - `Name` from aggregated week name
+      - `Weekly Analysis` from AI agent output
+      - `Type` as select value `Weekly Analysis`
+      - `Updated` as date
+    - Connect **AI Agent Weekly Performance Summary → Create Weekly Performance Card**.
+
+11. **Create the second Notion database schema**.
+    - Database name can be `AI Analysis`.
+    - Required properties and types:
+      - `Name` → Title
+      - `Weekly Analysis` → Rich text
+      - `Type` → Select
+      - `Updated` → Date
+    - Add select options including:
+      - `Weekly Analysis`
+      - `Global Summary`
+
+12. **Add a second Code node** for the global prompt.
+    - Rename it to **Prepare Global Summary Prompt with Indicators**.
+    - Configure it to:
+      - collect all weekly aggregated items
+      - sort them by `weekStart`
+      - build one text block summarising all weeks
+      - return a single item with a field called `globalPrompt`
+    - Connect **Aggregate by Week → Prepare Global Summary Prompt with Indicators**.
+
+13. **Add a second OpenAI Chat Model node**.
+    - Rename it to **OpenAI Chat Model Global**.
+    - Configure:
+      - Model: `gpt-4o-mini`
+      - Max Tokens: `600`
+      - Temperature: `0.3`
+    - Use valid OpenAI credentials.
+
+14. **Add a second AI Agent node** for the global summary.
+    - Rename it to **AI Agent Global Performance Summary**.
+    - Configure prompt mode as defined text.
+    - Use the incoming `globalPrompt` as the base prompt.
+    - Ask for:
+      - overall trend
+      - recurring bottleneck checkpoint
+      - best and worst weeks with reasons
+      - two recommendations
+      - short outlook
+    - Ask for bullet points only and no markdown headers.
+    - Use a system message focused on management-level OTIF analysis with British spelling.
+    - Connect:
+      - **Prepare Global Summary Prompt with Indicators → AI Agent Global Performance Summary**
+      - **OpenAI Chat Model Global → AI Agent Global Performance Summary** via AI language model connection
+
+15. **Add a final Notion node** to update the management summary card.
+    - Rename it to **Update Global Performance Summary**.
+    - Resource: `Database Page`
+    - Operation: `Update`
+    - Provide a fixed `Page ID` for the existing summary row.
+    - Map:
+      - `Name` = `Overall Performance Summary`
+      - `Weekly Analysis` = AI agent output
+      - `Type` = `Global Summary`
+      - `Updated` = current date
+    - Connect **AI Agent Global Performance Summary → Update Global Performance Summary**.
+
+16. **Create or identify the dedicated Notion page to update**.
+    - It must already exist in the analysis database or compatible target database.
+    - Copy its page ID into the node.
+    - Ensure the page contains compatible properties:
+      - `Name`
+      - `Weekly Analysis`
+      - `Type`
+      - `Updated`
+
+17. **Add credentials**.
+    - OpenAI credentials:
+      - assign to both OpenAI Chat Model nodes
+    - Notion credentials:
+      - assign to all three Notion nodes
+
+18. **Test the workflow with representative shipment data**.
+    - Confirm the source returns rows.
+    - Confirm the aggregation outputs one item per week.
+    - Confirm the weekly AI agent outputs a field named `output`.
+    - Confirm both weekly Notion pages are created.
+    - Confirm the global AI agent returns one summary.
+    - Confirm the final Notion page is updated.
+
+19. **Validate expression references carefully**.
+    - This workflow uses references like `$('Aggregate by Week').item.json.Name`.
+    - If you change branching structure or insert merge/filter nodes, item linking may need adjustment.
+
+20. **Add resilience improvements if rebuilding for production**.
+    - Add an IF node after aggregation to stop when `error = No shipments found`.
+    - Add deduplication logic in Notion if you do not want duplicate week entries.
+    - Optionally write AI weekly output back into the KPI database’s `AI Analysis` field.
+    - Consider limiting the global summary to a rolling date window if many weeks accumulate.
+
+---
+
+# 5. General Notes & Resources
+
+| Note Content | Context or Link |
+|---|---|
+| Supply Chain OTIF Performance Scorecard for Retail Logistics. Automatically aggregates shipment data, computes weekly OTIF KPIs, generates AI-powered performance analysis, and pushes everything to a Notion dashboard. | Workflow purpose |
+| Setup checklist: connect your data source to the shipment collection node; add OpenAI API keys to both model nodes; add Notion API credentials to all Notion nodes; verify Notion database IDs. | Operational setup |
+| Customisation ideas: adjust AI prompts to change analysis style or language; modify aggregation code to add custom KPIs. | Workflow maintenance and extension |
+| Video resource | https://www.youtube.com/watch?v=tOT8XhQ7eB8 |