Assortment Planning: Intelo.ai Agent Team

Assortment Planning is the Intelo Agent Team that takes you end-to-end through the assortment workflow (building the line list, clustering stores by demand, reconciling against OTB, and finalizing receipts) running continuously against your plan.

What is agentic assortment planning?

Agentic assortment planning is an AI operating model that owns the full workflow from line list through committed buy: item planning, demand-driven store clustering, OTB reconciliation, cluster/store-level eligibility, and buy planning. Intelo's Assortment Planning replaces four or five disconnected tools (and the spreadsheets between them) with one connected team that keeps the assortment reconciled with live demand and the financial plan.

Agents in this team

• Item Planning Agent: Builds the line list. Balances new references and capsule launches with proven carryover, surfaces AI recommendations, and persists merchant decisions back to the system.
• Store Clustering Agent: Groups boutiques and authorized retail doors by shared demand characteristics, not just geography. Produces cluster assignments that drive assortment depth and breadth.
• Assortment Reconciliation Agent: Fits unconstrained reference demand within OTB budget, space, movement availability, and inventory targets. Outputs a financially viable, capacity-feasible assortment plan.
• Cluster/Store Level Assortment Plan Agent: Assigns the reconciled reference list to specific store clusters/stores. Produces the definitive Reference × Cluster/Store eligibility matrix with the right depth and breadth per cluster/store.
• Buy Planning Agent: Finalizes purchase orders, configurations, and delivery schedules. Feeds committed receipts into the distribution network for downstream allocation.

Frequently asked questions

What does an agentic assortment plan actually output?

A connected set of artifacts: the line list (with new references, carryover, and capsule picks), the behavioral store clusters, the reconciled reference set that fits OTB and capacity, the Reference × Cluster/Store eligibility matrix, and the committed purchase orders with delivery schedules. Every artifact is produced by an agent, approved by a merchant, and persisted back to systems of record.

How does behavioral store clustering differ from traditional geography-based clustering?

Traditional clustering groups stores by region, zip code, or an analyst's manual tier. Behavioral clustering groups them by what customers actually bought, in what mix, at what sell-through rate. A boutique in Beverly Hills can cluster with one in Dubai if the demand pattern matches. The output is clusters that drive assortment depth and breadth decisions grounded in real behavior.

Can it integrate with our existing item planning workflow and PLM?

Yes. The Item Planning Agent reads from your PLM (ERP, Oracle Retail, SAP, Centric, custom) and persists merchant decisions back to it. The handoff to downstream buy planning uses standard formats so your purchasing and vendor-management workflows do not need to change.

How does it handle luxury and premium retail specifics (small batches, long lifecycles, tight margins)?

Intelo's product assumptions were built on luxury and specialty retail: small batches, cluster sensitivity, long product lifecycles, tight margin discipline. The clustering agent respects boutique-level behavior rather than averaging across the network. The reconciliation agent respects fabric-level and capacity constraints. The buy planning agent respects vendor minimums and delivery windows that luxury Houses actually face.

How long does it take to deploy Assortment Planning?

One quarter for the first team is typical across Intelo deployments. Assortment Planning is often the third or fourth team on, after Merchant Analytics and MFP, so the plan can be built against clean demand and inside a reconciled OTB. Expansion to adjacent teams happens without a new sales cycle.

How do merchants stay in control of the assortment decisions?

Every agent recommendation is explainable and overridable. The Item Planning Agent surfaces its suggestions; the merchant accepts, edits, or rejects in plain language. Overrides persist as signals the agents learn from. The output is the merchant's assortment plan, built with the agents' help, not an auto-generated plan the merchant has to re-check line by line.