Challenge
Every order at Avirago triggered a chain of manual tasks: creating purchase orders for suppliers, tracking deliveries, generating stock documents, converting currencies, and preparing shipments. With dozens of orders per week, the team was spending hours each day on repetitive data entry and document preparation.
These tasks were not only time-consuming but also highly prone to human error, especially when dealing with multi-currency documents and cross-system data lookups where a single mismatched record could cascade into bigger issues.
Solution
We analyzed Avirago's day-to-day operations, identified the biggest bottlenecks, and built seven n8n workflows tailored to their processes:
- Automated currency conversion for strategic documents. Saves ~5 minutes per document across ~20 documents per week (~1.5 hours saved weekly).
- Inbound payment processing. Saves ~5 minutes per document across ~20 documents per week (~1.5 hours saved weekly).
- Supply orders generation. Saves ~15 minutes per order across ~20 orders per week (~5 hours saved weekly).
- Invoice reporting. Saves ~10 minutes per run, executed once a month.
- Background ecommerce synchronizer. Runs every 15 minutes, replacing 1 to 3 hours of manual sync each month.
- Inbound delivery stock document generator. Saves ~20 minutes per document across ~5 documents per month.
- Stock release generator. Saves ~40 minutes per release across ~5 releases per month.
In total, these workflows save Avirago over 40 hours of manual work every month and have eliminated the most error-prone steps in their order pipeline.
Example workflows
Below are four of the workflows we built for Avirago.
Sticky notes and internal documentation have been removed from the screenshots.
Stock release generator
This workflow produces easy-to-read release documents for warehouse staff. It is triggered from Avirago's internal Telegram chatbot. When a release is requested, the workflow pulls the release data from the CRM via API and generates a PDF that includes the product name, product photo, barcode, quantity to release, and the address data needed for shipment.
A dedicated AI agent handles one of the more complex steps: the CRM and the product database use different identifiers with no common sync key between them. The AI searches the product database using the identifier from the CRM and resolves the correct product record so the document is filled with accurate details.
Currency conversion
Avirago's CRM has no native way to recalculate documents between two currencies using a specific exchange rate from the Polish National Bank. This workflow fills that gap.
It is triggered from an internal Telegram chat group. The user sends the order document; the workflow picks it up, calculates the correct prices in the target currency using official exchange rates from the Polish National Bank, and writes the updated record back to the CRM via its API. Once complete, it sends a confirmation back via Telegram.
Inbound payment processor
This workflow takes the document that confirms a payment has been settled and parses it with LlamaParse. It then searches for the matching records in the company system, marks the payment as paid, and notifies the team about the settled payment. Once done, it triggers another workflow responsible for generating purchase orders. The workflow also handles country-specific tax edge cases that would otherwise require manual verification.
Supply orders generator
This is one of the most extensive workflows, handling numerous edge cases and connecting records from two unsynced databases using AI. It relies on Avirago's CRM, LlamaParse to process supplier quotation documents, AI calls to reconcile unrelated database records, translation, and generating the final document. After the supply order is generated, it is sent for approve/decline (human-in-the-loop). On approval, the order is automatically emailed to the specific supplier.
Despite its complexity, the workflow includes robust error handling and has achieved ~95% accuracy over time, saving the Avirago team significant hours every week.
Results
By automating the most repetitive and error-prone parts of their order pipeline, Avirago reclaimed over 40 hours of manual work every month. Tasks that previously required careful attention and cross-referencing between systems now run in seconds with minimal human involvement.
The team can focus on growing the business instead of chasing data across spreadsheets and databases. Importantly, these savings scale with the company: as order volume grows, so does the time saved, making the workflows more valuable over time. The workflows continue to run reliably and are actively maintained as Avirago's processes evolve.