Pattern 1: AI Agent
An AI agent uses the SIMCEL API as its “memory” for supply chain data. The agent receives a natural language question, fetches the relevant KPIs, and generates a structured answer or recommendation. Use case: “What is our EBIT margin gap between Committed and Optimistic scenarios for FY2025?” Stack example:- Agent: Claude claude-opus-4-5 or GPT-4o with function calling
- Auth: SIMCEL API key stored in agent secrets
- MCP:
developers.simcel.io/mcpfor self-documenting context
GET /pnl-summary, GET /demand-forecast
Pattern 2: ERP Integration
SIMCEL scenario outputs feed into your ERP (SAP, Oracle, Microsoft Dynamics) to update financial forecasts, procurement orders, or production plans — keeping your operational system in sync with your planning scenarios. Use case: Every Monday, pull the latest Committed scenario P&L and demand forecast, transform to SAP format, and push updated budget figures into SAP CO/PA. Stack example:- Scheduler: Azure Logic Apps or cron job
- Transform: Python pandas to map SIMCEL fields → SAP BAPI structure
- Push: SAP RFC / BAPI or REST API depending on version
GET /pnl-summary, GET /demand-forecast
Pattern 3: BI Dashboard
SIMCEL scenario data is pulled into a Business Intelligence tool (Power BI, Tableau, Looker) to build live, interactive dashboards that business users can explore — without needing access to the SIMCEL app itself. Use case: A Power BI dashboard showing Committed vs Actual scenario comparison for EBIT, service levels, and CO2 — refreshed daily, accessible to 50+ business users. Stack example:- Connector: Python script using
requests→ writes to a SQL or Parquet file - Refresh: Scheduled via Power BI gateway or Airflow DAG
- Visualisation: Power BI with scenario slicers and waterfall charts
GET /pnl-summary, GET /demand-forecast, GET /supply-performance
Pattern 4: Direct API Client
A developer or data scientist queries the SIMCEL API directly in a script, notebook, or application — no middleware needed. Ideal for ad-hoc analysis, model training, or lightweight integrations. Use case: A data scientist pulls 12 months of demand data across 3 scenarios into a Jupyter notebook to train a demand sensing model. Stack example:Pattern 5: Slack / Teams Bot
An internal chatbot lets business users ask supply chain questions in plain language from Slack or Microsoft Teams — no login to SIMCEL required. Use case: Finance or supply chain teams get instant answers like “What’s the service level in DC-PARIS this month?” directly in Slack — without switching tools. Stack example:- Bot framework: Slack Bolt for Python or Microsoft Bot Framework
- LLM: OpenAI
gpt-4owith function calling - Deployment: AWS Lambda or a lightweight Docker container
Choosing the right pattern
| Pattern | Best for | Complexity | Time to build |
|---|---|---|---|
| AI Agent | Ad-hoc analysis, executive Q&A | Low | 1–2 days |
| ERP Integration | Automated budget sync, procurement | High | 2–4 weeks |
| BI Dashboard | Broad business user access, reporting | Medium | 3–5 days |
| Direct API Client | Data science, notebooks, prototyping | Very low | Hours |
| Slack / Teams Bot | Internal self-service, team alerts | Medium | 3–5 days |
All patterns share the same SIMCEL API credentials and rate limits. For high-volume integrations like ERP sync running multiple times per day, consider the Growth or Enterprise plan. See Pricing.

