Skip to content
← All casesML Pipeline + S&OPEnterprise

IBP: Demand Forecasting

ML model accuracy 88→95%, planning cycle 12→4 days, ₽263–550M/year impact

Problem

What doesn't work

Freight transportation planning was manual: commercial directors manually determined volumes by clients and routes. Production plan accuracy by clients — 88%, by routes — 59%. Planning cycle — 12 days. Labor costs — ~180 person-days per year.

Solution

Architectural approach

Automated demand forecasting across the network (unconstrained demand) with ML models. Integration into S&OP process: scenario planning, freight base optimization. 3 MVP phases: forecast for PV → expansion to KR → all RPS regions. Corporate architect on the project team.

Challenges

What made it hard

Commercial directors had planned manually for years and didn't trust ML forecasts — deployment required running both approaches in parallel for 6 months until ML proved superiority. Route accuracy was 59% — the model performed worse than humans on this dimension until seasonal features were added. Board justification: the accuracy-to-revenue correlation (0.798) was the only argument that convinced.

Role

My role & contribution

CTO / Technical Director

Corporate architect on the project. Designed ML forecast integration with Optimizer and Navigator. Defined the 3-phase MVP strategy. Justified the business case to the Board (₽263-550M/year impact, accuracy-to-revenue correlation = 0.798).

Demo

How it looks

Architecture

System architecture

Historical Dataclients, routesvolumesML Modelsforecasting engineClient Forecast88 → 95%accuracyRoute Forecast59 → 75%accuracyS&OP PlanningscenariosOptimizerNavigatorPlanning Cycle: 12 → 4 daysCorrelation with marginal income = 0.798263–550M₽/year effectAI/LLMDataInfraEval
Implementation

How it works

ML models for client and route forecasting. Automated route volume calculation replacing manual process. Integration with Optimizer and Navigator. Daily freight forecast for scheduling. Prototype → MVP 1.0 (PV) → MVP 2.0 (PV+KR) → MVP 3.0 (all RPS).

Architecture Decision

Why this way

ML forecast + S&OP instead of pure BI automation

Alternative

Automate existing manual process without ML (dashboards + alerts)

Why it didn't fit

BI automation: speeds up the same manual process but doesn't improve accuracy. ML model: new forecast quality that humans can't achieve manually. Accuracy-to-revenue correlation = 0.798.

Result

Accuracy +7 pp, cycle 3× faster, ₽263–550M/year impact. Foundation for company's S&OP process

Metrics

Results

01
Client accuracy: 88% → 95%
02
Route accuracy: 59% → 75%
03
Planning cycle: 12 days → 4 days
04
Freed 779 person-days/year (~3 FTE)
05
₽263–550M/year total impact
Business Impact

Impact on business

Three impact areas: S&OP scenarios (₽150–300M/year from correct obligation determination), marginal revenue increase (accuracy-to-revenue correlation = 0.798), 5–10% reduction in over-provisioning (₽113–250M/year). Total: ₽263–550M/year. Freed 779 person-days for commercial directors.

Methods

Algorithms & patterns

ML Forecasting (demand)S&OP PlanningScenario AnalysisCorrelation AnalysisA/B Testing (plan accuracy)
Stack

Technologies

  • Python
  • ML Pipeline
  • Anaplan
  • SAP BW
  • S&OP Framework

Ready to discuss?

If you need an architect who builds autonomous AI systems — reach out.

Serbia-based · CET/CEST timezone · EU-aligned working hours · International contracts experience