WENDA AI Invoice Extraction
AI product management for a B2B logistics SaaS startup in Italy
At WENDA, an Italian B2B SaaS startup in supply chain and logistics, I worked on an AI module that extracted structured information from invoices and turned it into usable operational data.
Status: Delivered as part of product development / PoC work
Type: B2B SaaS / AI Product / Logistics
Role: AI Product Manager
Focus: Product discovery, requirements definition, UX design, cross-functional execution, client-facing PoC
Why this project matters
Operational data is often trapped inside business documents
In logistics and supply chain operations, critical information is often embedded in invoices and other business documents. Teams need that information to be accurate, structured, and usable inside operational workflows.
The challenge was not simply extracting text. The product needed to turn document information into reliable operational data that could support real business processes.
What I discovered
AI value depends on workflow fit, not just extraction accuracy
Working with executives, customer success, engineers, and designers, I saw that AI document extraction only becomes valuable when it fits into the user’s operational workflow.
Users do not just need extracted fields. They need confidence, reviewability, error handling, and a clear path from raw documents to business actions.
Product challenge
Turning unstructured documents into usable operational data
The product challenge was to design an AI-powered module that could extract structured information from invoices and make it usable within a B2B logistics SaaS product.
This required balancing user needs, technical feasibility, data quality, UX clarity, and client-specific operational requirements.
What I delivered
Requirements, UX, and PoC-ready product direction
I led product discovery, requirements definition, and UX design for the AI-powered invoice data extraction module.
I worked closely with the CEO, COO, customer success, engineers, and designers to translate ambiguous business needs into product scope, user flows, specifications, and a direction that could support client-facing PoC work.
Evidence of execution
AI product work inside a real B2B SaaS environment
This project included AI product discovery, invoice data extraction workflow design, requirements definition, UX design, cross-functional coordination, and client-facing PoC support.
It also gave me hands-on experience operating close to leadership in an Italian startup, connecting market needs, product strategy, technical implementation, and enterprise client expectations.
What this demonstrates
Real-world AI product management and B2B execution
This project demonstrates my ability to manage AI product work in a real B2B SaaS environment, not only as a concept.
It is evidence of AI product management, requirements definition, workflow thinking, UX design, PoC execution, and cross-functional collaboration with executives, engineers, designers, and customer-facing teams.
