n8n + WhatsApp AI Automation for Parcel Delivery Operations
Budget: $250 – $750 USD
I need a production-ready automation that mirrors the daily routine of a parcel delivery depot. Couriers send a photo of their paper manifest to a dedicated WhatsApp number; from there everything should flow without human touch until a final, concise review screen.
Workflow overview
• n8n listens to incoming WhatsApp messages (using the Cloud API or an approved gateway).
• The attached photo—text only, no barcodes or QR codes—is routed to an AI extraction step (OpenAI Vision, Google Vision, or a model you propose).
• Parsed order data lands in a staging sheet inside Google Sheets.
• An automated validation layer cross-checks obvious errors; anything flagged is pushed back to me inside WhatsApp for a quick thumbs-up or correction.
• Once confirmed, the data is appended to the live operations sheet and a status message returns to the courier so they know the manifest was accepted.
Key expectations
– Both core tasks must be covered: precise data extraction from the photos and automated, conversational WhatsApp messaging.
– The flow has to be resilient enough for real-world volume; every weekday several dozen manifests will arrive almost simultaneously.
– Logging, retries, and error notifications should be baked in so we discover issues before the drivers do.
– Please keep the solution focused—this is an operational tool, not a polished SaaS product.
Deliverables
1. n8n workflow export(s) with clearly named nodes and comments.
2. Google Sheets template with staging and live tabs, all formulas or Apps Script needed for data hygiene.
3. Read-me or short Loom walkthrough explaining setup, WhatsApp credentials, and how to tweak the AI model thresholds.
4. One week of light post-launch support to iron out edge cases.
If you have prior logistics or last-mile experience, mention it—real-world insight is more valuable than shiny dashboards.
Workflow overview
• n8n listens to incoming WhatsApp messages (using the Cloud API or an approved gateway).
• The attached photo—text only, no barcodes or QR codes—is routed to an AI extraction step (OpenAI Vision, Google Vision, or a model you propose).
• Parsed order data lands in a staging sheet inside Google Sheets.
• An automated validation layer cross-checks obvious errors; anything flagged is pushed back to me inside WhatsApp for a quick thumbs-up or correction.
• Once confirmed, the data is appended to the live operations sheet and a status message returns to the courier so they know the manifest was accepted.
Key expectations
– Both core tasks must be covered: precise data extraction from the photos and automated, conversational WhatsApp messaging.
– The flow has to be resilient enough for real-world volume; every weekday several dozen manifests will arrive almost simultaneously.
– Logging, retries, and error notifications should be baked in so we discover issues before the drivers do.
– Please keep the solution focused—this is an operational tool, not a polished SaaS product.
Deliverables
1. n8n workflow export(s) with clearly named nodes and comments.
2. Google Sheets template with staging and live tabs, all formulas or Apps Script needed for data hygiene.
3. Read-me or short Loom walkthrough explaining setup, WhatsApp credentials, and how to tweak the AI model thresholds.
4. One week of light post-launch support to iron out edge cases.
If you have prior logistics or last-mile experience, mention it—real-world insight is more valuable than shiny dashboards.