Skip to main content

What is AInventory

AInventory is Directrix's platform for collaborative and intelligent demand and inventory planning, built for Supply Chain teams. It brings together statistical forecasting, cross-functional business collaboration, and inventory optimization in a single environment, under Directrix's proprietary Inventory Evolution System™ framework.

The predict-then-optimize approach

AInventory follows a predict-then-optimize logic: first, it builds the best possible forecast through a multi-phase collaborative cycle; then it feeds that forecast into an inventory optimization engine to compute reorder points, safety stock levels, and expected fill rate. This way, forecasting errors propagate to inventory decisions in a controlled and measurable manner.

Module 1 — Collaborative Forecast

The Collaborative Forecast module organizes demand forecasting across four sequential phases:

PhasePrimary actorDescription
Baseline (AI)Statistical engine / SofIAAutomatic generation of the initial forecast using artificial intelligence models.
SalesCommercial teamBaseline adjustment with market intelligence, commercial agreements, and field signals.
MarketingMarketing teamEnrichment with trade plans, campaigns, and promotional activities.
ConsensusSupply Chain leadershipFinal version agreed upon across Supply Chain, Sales, and Marketing; used as the operational input.

Each phase generates a recorded version of the forecast. The platform measures forecast quality through KPIs (per SKU, per version, and cumulative) and applies FVA (Forecast Value Added) to quantify how much value each phase adds over the previous one.

Module 2 — Inventory Optimization

The Inventory Optimization module takes the Consensus forecast and per-SKU parameters and uses Monte Carlo simulation to compute optimal replenishment policies. The key outputs are:

  • Factor k — number of safety standard deviations that minimizes expected total cost.
  • Fill rate — probability of meeting demand without a stockout.
  • Reorder point and safety stock level.

Cross-cutting element — SofIA

SofIA is AInventory's AI copilot. It is available from any screen via the floating button in the bottom-right corner. SofIA answers questions about tenant data, explains KPIs, interprets simulation results, and guides users through platform features.

Multi-tenant architecture and roles

AInventory is multi-tenant: each company (tenant) operates in an isolated data space. Within each tenant, roles (Administrator, User, Owner) control which modules and actions are available to each person.

Where to start?

If this is your first time in AInventory, read Key concepts in 5 minutes next to learn the essential vocabulary. Then choose your path:

  • Forecast track → Tutorial: full collaborative cycle.
  • Inventory track → Tutorial: simulation and parameters.