What's new?
The AI agent doesn't just explain what's on screen. It traces every forecast and every supply request back to its foundations: which historical data was used, which forecasting models were considered, why one of them won out, and which business rules factored into the final result. Planners and buying teams simply ask their question in the chat next to the relevant view — no exports, no waiting on an analyst.
An example from planning:
A planner notices a sharp increase in the forecast for a component and asks: "Why is the forecast for this product higher than before?" The agent shows the historical data used, whether the demand pattern has changed, which methods were evaluated — and why the chosen method performed best.
An example from purchasing:
As a purchasing lead, want to understand a supplier request? Ask: "Why are we requesting 240 units from this supplier?" The agent traces the quantity back to the forecast, current stock, open orders, safety stock, lead time, and order constraints — ready to double-check before the request goes out, or to explain to management.
Where to find the AI-Agent?
-
In "Products"
Navigate to "Products" in the menu bar and open any desired SKU. "PartsOS" will then appear on the right, and clicking it opens the chat.
-
In "supply requests"
Open any procurement request and click on "PartsOS" on the right — the chat opens right next to it.
How it works
Open any supply request and click "PartsOS" on the right — the chat opens right next to it. Suggested questions help you get started; your question is checked against the PartsOS system (visible while data is being retrieved), and the answer is structured clearly: a summary first, then the reasoning in detail, and, where relevant, a concrete recommendation with quantities and dates.
Why it matters
Automated recommendations only get used once people understand how they came about. The AI agent makes that path traceable, without pretending that every model decision can be reduced to a simple sentence.
The result: faster review, more transparency — and more trust in automated planning. PartsOS doesn't just deliver a recommendation. It shows how the system arrived at it.
FAQs
-
What can the PartsOS AI agent do?
It explains every forecast and supply request in plain language — tracing back to the data used, the forecasting model chosen, and the calculation steps behind it.
-
Do I need to phrase my question in a specific way?
No — the agent understands plain, everyday questions, not commands or codes.
-
What data does the agent rely on?
Verifiable, planning-relevant data: historical demand, forecasting models and how they were selected, stock levels, open orders, lead times, safety stock, and configured business rules.
-
Does the agent replace my own review?
No, it makes your review faster and better informed. The decision to accept, adjust, or reject a recommendation stays with you.