--- title: "Trusted Trade Agents – Junction 2026" description: "How can a food-service wholesaler let autonomous purchasing agents place an order – finding products, comparing offers and getting the right approval in a fraction of the time it takes today – without exposing real systems, confidential data or the marketplace roadmap?" url: https://junction2026.com/challenges/valio-aimo-trusted-trade-agents --- > For the index of all pages on this site, see [llms.txt](https://junction2026.com/llms.txt). # Trusted Trade Agents Junction 2026 · Agentic AI Track 105 teams by Valio Aimo How can a food-service wholesaler let autonomous purchasing agents place an order – finding products, comparing offers and getting the right approval in a fraction of the time it takes today – without exposing real systems, confidential data or the marketplace roadmap? ### Introduction Valio Aimo® serves 20,000+ professional kitchens and is planning to build a digital B2B marketplace connecting foodservice buyers and suppliers. As ordering shifts from people clicking through a catalog to **AI agents making autonomous purchasing decisions**, the question of how a wholesaler exposes itself to those agents – what they can see, decide and be trusted with – becomes existential for data ownership, trust, and who owns the customer relationship. The Digital B2B Marketplace is a future step in this direction; for Junction 2026 we are mocking today's webshop and backend systems, not the marketplace itself. We want you to prototype the **interface and operating model** that would let an external purchasing agent interact with Valio Aimo's systems, built on **open, verifiable standards** rather than a single platform's black box. ### The challenge B2B ordering in foodservice and wholesale is still largely manual or menu-driven: a buyer logs into today's Valio Aimo webshop and searches a catalog. The step we want to explore is **external purchasing agents** that find, compare, order and handle exceptions on the buyer's behalf, working with Valio Aimo directly – not by building a whole marketplace, but by defining the interface and operating model through which an external agent could interact with Valio Aimo's systems at all. #### Core – your solution must enable - **Product, price & availability lookup:** a structured way for an external agent to search for and retrieve what's available and at what price. - **Requesting price & availability, then placing an order:** a defined interaction for moving from inquiry to commitment, based on fixed prices rather than a negotiated quote. - **Permission & human-approval management:** what the agent may decide on its own, and what must go to a human for sign-off. #### Optional – open directions (not required to win) - **Delivery time & sustainability comparison:** not just price, but delivery reliability and the certificates/sustainability labels already shown in the webshop. _(Category-level carbon-footprint ranges are a stretch goal.)_ - **Substitution and other exception handling:** what happens when the exact request can't be met. _(Last year's Zero-Fail Logistics challenge covered failed deliveries in depth, so this isn't scored as core.)_ - **Traceability and explainability:** a human-readable record of what the agent did and why. #### Example request > _"I need 1,000 litres of lactose-free milk by Monday. Find an option that meets our requirements based on price and availability."_ A hotel chain's agent sends this request. Your solution returns products, prices and availability in structured form; the agent places the order within its authority or sends it to a human for approval. **Impact:** This is not about replacing today's e-commerce UI with a chatbot. It is about defining how a European food-service wholesaler can be safely and openly readable by autonomous agents – before that interface gets defined by someone else's platform. ### Valio Aimo® Valio Aimo® was created in 2021 as a bold leap forward by Valio, transforming traditional wholesale into a next-generation foodservice powerhouse. We are more than a distributor: Aimo is where the breadth and agility of wholesale meet the deep product expertise, R\&D capability, and brand-building strength of a global food innovator. - **20,000 professional customers** across restaurants, foodservice operators, and industrial customers rely on us to keep their business running every single day. - **25,000 product names** sourced from **700 trusted suppliers**, forming one of the most comprehensive foodservice assortments in Finland. - **350 specialists** are committed to delivering excellence in logistics, service, and digital transformation. - **Part of Valio**, Finland's leading dairy and food company, with:   - **€2.4B in revenue**   - Operations in **60 export markets**   - A pioneering sustainability mission: **carbon-neutral milk by 2035** As part of Valio, Aimo represents the cutting edge of how foodservice can be reinvented for the AI-driven age. **Valio Aimo® is not just serving today's kitchens; we are shaping the future of foodservice.** ## Insight ### Problems we'd love to see tackled - **Trust protocol:** the minimum data/verification standard for an external agent to trust Valio Aimo's claims on stock, price and origin without a manual check. - **Guardrails model:** which decisions require human sign-off (spend threshold, substitutions outside stated preferences) vs. fully autonomous. - **Response logic:** how an agent handles partial availability and substitution ranking ("good-better-best") within Valio Aimo's own assortment. - **Audit trail:** a usable, human-readable log of an agent's purchasing decisions, for both buyer and Valio Aimo. ### Example scenarios to spark ideas - **Need → recipe → order:** a chef plans next week's menu around a seasonal ingredient. The agent suggests recipes, works out what's missing from stock, and places the order. - **Last-minute event:** an unexpected booking for tomorrow. The agent identifies what the recipes require, flags what can't be sourced in time, and proposes substitutions through recipe logic rather than a simple SKU swap. - **Supplier stock-out:** a key product runs out mid-week. The agent proposes a recipe-compatible substitute before the kitchen notices the gap, and explains the trade-off. - **Proactive selling agent:** Valio Aimo's own agent spots a new or seasonal product and proactively suggests it to the buyer's agent. - **Allergen-safe substitution under pressure:** in a school or hospital kitchen, a substitution must still satisfy strict allergen and nutrition rules. - **Standing order drift detection:** the agent learns a kitchen's recurring weekly order and flags anomalies (portion counts creeping up, cost drifting) before auto-renewing it. **Scope note:** we are mocking a structured data/API interface. Whether Valio Aimo's production interface eventually becomes a traditional API, an MCP-style agentic interface or a full service agent of its own is outside the scope of this challenge. ## Resources ### What we'll bring - **Synthetic product catalog, simulated suppliers and mock APIs** mimicking Valio Aimo's webshop/backend systems, built on realistic data: prices, availability behaviour, delivery lead times, and sustainability certificates/labels (category-level carbon-footprint ranges as a stretch goal). - **No real-time, live or confidential data or systems** – everything you build against is synthetic. - **Mentoring** on AI agents, wholesale trading rules, exception handling and data governance, plus coaching on which background data and rules are essential. **Preferred tech:** AI agents (agentic/tool-use frameworks), open data/interop standards, recommendation logic, optional verifiable-credential or trust-layer approaches. Agentic payment standards (e.g. x402, Nevermined) are relevant context but not required. ### Mentors - **Marketplace/Commercial:** Hanna Forsström, Pirjo Kaartinen - **Online Sales:** Minni Ruhtinas - **IT/Tech integrations & Data governance:** Tommi Nurmi-Aro, Ville Takanen, Riku Tapper **Access the challenge resources here:**\ **Resource portal password:** `aimo_ai_challenge` The resource portal contains downloadable datasets, API documentation and instructions for accessing the mock backend systems. **API authentication:** API key: `junct_tDnQn4VI_nfhELGCOEkY4vbuxm324MmYmILuokYG03Y` Use this API key to authenticate requests to Valio Aimo's mock APIs. Refer to the documentation in the resource portal for integration instructions. **Webshop test accounts:** Test credentials and access instructions will be shared separately once available. All provided data is synthetic. No production systems or confidential business data are involved. ## Challenge partner Valio Aimo Agentic AI Track ## Company Info: Valio Aimo Valio Aimo® was created in 2021 as a bold leap forward by Valio—transforming traditional wholesale into a next-generation foodservice powerhouse. We are more than a distributor: Aimo is where the breadth and agility of wholesale meet the deep product expertise, R\&D capability, and brand-building strength of a global food innovator. 20,000 professional customers across restaurants, foodservice operators, and industrial customers rely on us to keep their business running every single day. 25,000 product names sourced from 700 trusted suppliers, forming one of the most comprehensive foodservice assortments in Finland 350 specialists are committed to delivering excellence in logistics, service, and digital transformation. As part of Valio—Finland’s leading dairy and food company with €2.3B in revenue, operations in 50 export markets, and a pioneering sustainability mission (carbon-neutral milk by 2035)—Aimo represents the cutting edge of how foodservice can be reinvented for the AI-driven age. Valio Aimo® is not just serving today’s kitchens; we are shaping the future of foodservice. Website: ## Judging Criteria - Working end-to-end demo — (30%) The solution runs through a full request → lookup → order/approval flow in the mocked environment, not just isolated pieces. - Solving the problem and user need — (20%) how Precisely the agent matches what the buyer actually asked for, with the least manual correction needed. - Autonomy and controls — (20%) Quality of the guardrails design. What the agent can decide alone, what needs human sign-off, and how decisions are logged and auditable. - Robustness to exception situations — (15%) Handling of partial availability, substitutions and delivery constraints beyond the happy path. - Openness and credible path toward production — (15%) Open, documented standards rather than a proprietary format, clear visibility of assumptions, well-justified technical choices, and a credible view of next development steps ## Prizes - 1st place 4 000€ - 2nd place 2 000€ - 3rd place 1 000€