Azirella
← Back to Autonomy

Domains

Demand & Supply Planning

End-to-end planning from demand forecasting through supply plan generation, built on the AWS Supply Chain 3-step process and executed by agents under AIIO. Two of the six decision domains Autonomy covers.

"By 2026, 75% of large enterprises will have adopted AI-augmented planning... moving from deterministic to probabilistic planning."

, Tim Payne, VP Analyst, Gartner

"The biggest opportunity in supply chain planning isn't better algorithms, it's closing the gap between what the plan says and what actually happens on the ground."

, Knut Alicke, Partner, McKinsey (McKinsey, "Autonomous supply chain planning for consumer goods companies," 2024)

The AWS SC 3-Step Planning Process

1. Demand Processing Aggregate & time-phase demand 2. Inventory Targets Safety stock & policy calc 3. Net Requirements Netting, BOM explosion & sourcing
75%

of enterprises adopting AI planning by 2026

Gartner

30-50%

reduction in lost sales from better demand sensing

McKinsey

20-30%

reduction in inventory carrying costs

BCG

$2.8T

global supply chain planning market by 2030

Grand View Research

Step 1: Demand Processing

Aggregate demand from statistical forecasts, customer orders, and consensus inputs. Net out committed and allocated inventory, then time-phase demand across the planning horizon with P10/P50/P90 percentiles for uncertainty quantification.

Step 2: Inventory Target Calculation

Calculate safety stock and target inventory levels using four policy types:

  • Absolute Level - Fixed quantity targets (simplest)
  • Days of Coverage (Demand) - Dynamic targets based on actual demand
  • Days of Coverage (Forecast) - Dynamic targets based on forecast
  • Service Level - Statistical safety stock with z-score calculation

Hierarchical overrides ensure the right policy applies at the right level: Product-Site overrides Product overrides Site overrides Config defaults.

Absolute Level Fixed quantity targets Days of Coverage (Demand) Based on actual demand Days of Coverage (Forecast) Based on forecast data Service Level Statistical safety stock with z-score calc Simple Sophisticated

Step 3: Net Requirements Calculation

Time-phased netting computes gross requirements minus on-hand minus scheduled receipts. Multi-level BOM explosion handles recursive component requirements. Sourcing rules with priorities determine whether to buy, transfer, or manufacture. Lead time offsetting ensures orders are placed when they need to be, not when they're reviewed.

Gross Requirements - On-hand Inventory - Scheduled Receipts = Net Requirements Buy Transfer Manufacture

Master Production Scheduling (MPS)

Strategic production planning with rough-cut capacity checks. MPS drives finished goods production at the aggregate level, balancing demand requirements against available capacity and material constraints.

"The shift from deterministic MPS to probabilistic planning is the most significant change in production planning in 30 years. Organizations that master this transition will have a fundamental competitive advantage."

, Kaitlynn Sommers, Senior Director Analyst, Gartner Supply Chain

"By interacting with planners, we get all of their knowledge and experience. The planner's implicit knowledge can be made into a standardized, explicit process."

, Knut Alicke, Partner, McKinsey & Company (McKinsey, "Beyond automation: How gen AI is reshaping supply chains," April 2025)

Material Requirements Planning (MRP)

Detailed component requirements exploded from MPS. Multi-level BOM processing with scrap rates, yield adjustments, and configurable lot sizing rules (LFL, FOQ, POQ, EOQ, MIN/MAX).

Capacity Planning

Resource utilization analysis identifies bottlenecks before they become problems. Rough-cut capacity planning at the MPS level and detailed CRP at the MRP level ensure plans are feasible before release.

Industry Perspective

"AI in planning works best when it augments human judgment, not when it replaces it. The most successful implementations keep planners in the loop."

, Sam Ransbotham, Professor, Boston College & MIT Sloan Management Review

"Companies using AI-driven planning see 10-20% improvement in forecast accuracy and up to 65% reduction in manual planning effort."

, Shervin Khodabandeh, Managing Director, BCG

"Only 7% of supply chains can sense and shape demand in real-time. Autonomous planning closes this gap."

, Lora Cecere, Founder, Supply Chain Insights
65%

reduction in manual planning effort

BCG, 2024

10-20%

improvement in forecast accuracy with AI

BCG/MIT Sloan

7%

of supply chains can sense demand in real-time

Supply Chain Insights

The S&OP Cycle

S&OP Cycle Demand Review Supply Review Pre-S&OP Executive S&OP

See planning in action

Walk through the full planning cycle from demand signal to supply plan execution.