AGR

Our customer survey found that 82% of planners require clear logic before they will trust an AI output. That result didn’t surprise me. If anything, it confirmed something that has been core to how we develop at AGR for a long time: transparency has to be built into the product, not added afterwards. It becomes even more important with AI. When we designed AI Sales History, we were asking planners to trust something quite complex: building a sales history for an item based on the behaviour of comparable items. We could simply have surfaced the result. Instead, we show how we got there. Planners can see which comparable items were selected, why they were selected, the weightings applied and how the resulting sales history was constructed. They can question the matches, adjust the inputs if they disagree, and every override is logged so the reasoning remains traceable. That matters because this isn’t just an interesting AI recommendation. The sales history feeds directly into things like safety stock, reorder points and opening order quantities. If a planner can’t understand where the numbers came from, it’s very difficult for them to confidently own the decisions that follow. For me, this is one of the most important principles as we bring more AI into inventory planning. Transparency isn’t a feature of our AI strategy. It’s folded into how we build the product. 👉 https://lnkd.in/eNsHgh2a

Your Microsoft ERP records what happened. AGR helps you decide what to do next with better planning.

Extend your Microsoft ERP with advanced forecasting, inventory optimisation and automated replenishment.

Microsoft ERP data becomes actionable with automated forecasting, replenishment and inventory insights.

A retailer launching 200 new lines ahead of the Christmas range has a planning problem most systems quietly ignore. No sales history. No baseline. Just a blank field where the forecast should be. The usual fix is a manual estimate - copy a similar product, apply a gut-feel adjustment, and move on. It works for ten items. At 200, it becomes the single biggest source of forecast error in the business, and nobody knows until the markdowns start. New Item Forecast builds that opening estimate from comparable items that have actually sold - matching on attributes, price point, category, and demand behaviour. The planner gets a starting point grounded in real data rather than a copied cell. We asked planners what they needed from AI before they would trust it. 80% said they required the ability to review and override recommendations before acting. New Item Forecast is designed around exactly that - the AI produces the starting forecast, the planner examines it, adjusts it, and owns the decision. What changes is not who decides. It is what they are applying judgement to.👉 https://lnkd.in/e6G6XkqJ

A buying team at a mid-sized wholesale distributor runs a weekly exception report. On a busy week, it has 300 items on it. They look at maybe forty. Not because they are not trying. Because properly reviewing an item - pulling the forecast, checking the lead time, looking at the stock position, understanding why it was flagged - takes fifteen to twenty minutes. At 300 items, that is not a review. It is a selection process, and the criteria for selection are usually proximity to crisis. There is a second problem inside this one. Even the forty items that do get reviewed are often reviewed superficially, because the planner does not have time to understand why the flag was raised - they just act on the fact that it was. 82% of planners say they will not trust an AI output without a clear explanation of the logic behind it. If the flag does not come with reasoning, the review is not really a review. Item Advisor fixes both. One click per item gives the planner forecast logic, risk, ABC classification, lead time position, and item health, written in plain language with a recommendation. The review that took twenty minutes takes two - and it is an actual review, not a rubber stamp. The exception list does not get shorter. But the proportion of it that gets proper attention does. 👉 https://lnkd.in/evtXDZR9