Field2Fork AI Hub (Apulia)

Field2Fork AI Hub: a 'Personal Farmer' for Foggia

main tellingexample world: italydepth: standard2 min markdown

Field2Fork AI Hub came out of the third rung of the PoliRural+ open calls in the province of Foggia, Apulia. Two earlier projects prepared the ground.

Before the build: listen, then co-create

• DEMO (Outreach call, Cluster Srl) mapped needs: a producer survey, an online consultation on 26 February 2025 and five focus groups. It proposed cooperative electric or refrigerated vehicle pooling through Local Action Groups to tackle logistics and the low level of cooperation between entrepreneurs.
• INCO (Mobilise call, Virgolettere di Rocco Devito) ran a multi-actor hackathon that produced four prototypes. The winner, RuralTech Innovators’ Short Supply Chain Anti-Waste System, matches consumer pre-orders with producers’ seasonal margins and sends surplus or off-grade produce to local markets at a discount. The teams also tested PoliRural+ components such as UseAdvisor and JackDaw.

What Field2Fork built

EXO Soc. Cons. a r.l., with Distretto Agroalimentare Regionale S.C.R.L., won a Develop-call grant to link small producers with urban buyers. The team collected product information from local producers into a first digital catalogue and built a dual-track marketplace:

• a conventional, search-optimised catalogue store;
• the “Personal Farmer”, a conversational assistant. The pilot website says it “interprets natural-language dietary and ethical preferences to generate multiple, personalized shopping-cart scenarios”;
• a zero-kilometre logistics engine that matches the buyer’s location with local farm pickup hubs, so orders are collected nearby instead of shipped.

The team promoted the platform through regional media (Foglie, L’Attacco), workshops and a webinar.

What came out

Two user-acceptance surveys reached 170 consumers and 23 agri-food producers. Consumers rated usability 4.44/5 and the Personal Farmer’s information clarity 4.64/5, and 90.3% gave positive evaluations. Producers rated the platform’s value 4.25/5.

These numbers measure acceptance, not market impact. The D6.3 summary reports no sales volumes or changes in farm income. Whether the platform moves Apulia towards its target of +30% short-chain volume by 2028 still has to be measured.

The Apulia sequence of listening (DEMO), co-creating (INCO) and then building (Field2Fork) is the open-call ladder in miniature. Results across all calls are compared in open-call results.

Inside the Personal Farmer: prompts, intents and pickup hubs

another tellingexample world: italydepth: technical2 min markdown

D6.3 describes the Field2Fork AI Hub as a containerised, hybrid e-commerce microservices architecture: a PrestaShop back end with a separate conversational front end. The split matters more than any single model.

Four moves from chat to order

1. Structured system prompting. The assistant works under a structured system prompt that frames the conversation around the producers’ catalogue.
2. Intent routing. Each message is routed by what the user intends to do, so that a question about a product, a basket request or a pickup request reaches the right handler.
3. Dynamic HTML cards. Product suggestions appear as cards rendered from catalogue records, so the user sees catalogue items with their shop data rather than free-text descriptions.
4. Named-entity recognition (NER) for logistics, then a tokenised checkout. NER extracts the place a user mentions, and the zero-kilometre engine matches it to the nearest farm pickup hub. A tokenised checkout URL then hands the basket to the ordinary shop checkout.

In other words, the shop remains the system of record: products, prices, orders and payment stay in PrestaShop. The AI layer translates between how people talk (“something vegetarian for six, no nuts”) and what the catalogue holds. The same catalogue also serves a conventional, search-optimised store, which is the second track for users who prefer not to chat.

Retrieval is another option

The old version of this block described Field2Fork as a retrieval-augmented generation (RAG) system. D6.3 does not say that. Within the Develop call, RAG is documented for DASK on Kythera: its localised JackDaw chatbot uses RAG over a knowledge base, with voice input for Greek-language farming advice. LocalLinks.AI uses AI differently again, for semi-automated data extraction with confidence scoring when it builds its producer database.

The hidden cost

In all three cases someone has to curate the data. The cross-pilot policy brief for the Commission (D7.3, Annex 10) found that feeding rural, multilingual data into PoliRural+ AI systems “required months of manual work: annotation, formatting, and validation”. It recommends a “Data Curation Line” in EU AI project budgets. A Personal Farmer is only as good as the catalogue the producers keep up to date.