The observation
Italian SMEs already have access to AI tools. Several interview participants used tools such as GitHub Copilot, Claude or ChatGPT individually, without shared standards, governance or a way to measure what was working. The problem was not tool availability. It was adoption without coordination.
I ran 30+ structured interviews with Italian SME owners and operations managers across 2025 and 2026. A recurring pattern was that participants did not want “AI” as an abstract capability. They wanted specific operational problems solved, early evidence of value and limited infrastructure overhead.
What I built, and how
Not as a PM writing specs. I designed the API architecture before writing a line of implementation code, configured the retrieval engine, built and tested the agentic workflows on n8n, integrated Claude API as the main LLM with governance constraints, and validated the endpoints directly, including error handling and edge case behavior.
The product lets a small company load its operational documents: contracts, procedures, internal knowledge, recurring workflows. The system indexes them adaptively, builds a conceptual graph connecting entities and concepts across documents, and lets people query that knowledge base. The response path is designed to use retrieved evidence and return an insufficient-context response when the source material does not support an answer.
The retrieval architecture: AREX
The core engine is AREX (Adaptive Retrieval for Extended Experience), an advanced RAG architecture developed by researchers Samuele Pretini and Giorgia Lentoni (Milan, 2025). I adopted it as the technical foundation and applied it to the Italian SME context.
A standard RAG system has three structural problems: the schema is decided before the content is known, there is no memory of past interactions, and the LLM can still generate answers not anchored to the retrieved evidence. AREX addresses all three.
Metadata is induced dynamically at ingestion time, adapting to each document's actual structure rather than forcing it into a preset schema. Retrieval runs in parallel across four channels: keyword search on metadata and text, sparse lexical search for exact matches, dense semantic search on vector embeddings, and graph traversal on a conceptual graph connecting entities and concepts across documents. A reasoning agent integrates the results using retrieved evidence. An interaction experience graph records queries, responses and corrections so later retrieval and workflow decisions can be improved without retraining the LLM; improvement is not automatic or quantified here.
What the interviews taught me about SME adoption
Recurring concerns included making a business decision from a plausible but wrong answer, the onboarding burden of another tool and data sovereignty.
Those three fears shaped the product directly. Grounded-only output with source citations handles the first. A no-code document loading interface handles the second. A self-hosted deployment option handles the third.
Results
- Working MVP/pilot used to test the architecture and product assumptions
- 30+ structured interviews conducted before and during build
- Common automation and enterprise AI platforms benchmarked to test the product assumptions against existing options