Lokad Quantitative Supply Chain Platform

Lokad was created in Paris in 2008 as a software company with a mathematical core. It addresses stock-outs, write-offs and manual planning work by automating daily supply chain decisions for retail networks, aerospace, automotive, fashion, oil and gas and manufacturing. It has grown organically, funded by revenue rather than investors.

The Lokad platform combines probabilistic forecasting, differentiable programming and its own Envision language, released in 2014, to compute purchasing, allocation, dispatch, assortment, scheduling, pricing and stock decisions ranked by financial impact. Dedicated Supply Chain Scientists build and run each customer's numerical recipe on top of data extracted from ERP and other systems. Customers include Air France, Worten, Trek and Mister Auto.

• Probabilistic forecasting: quantile and probabilistic demand and lead-time forecasts.
• Purchasing and sourcing: reorder quantities ranked by expected financial return.
• Allocation and dispatch: multi-echelon stock allocation across DCs and stores.
• Pricing and promotion: algorithmic pricing across large catalogs.
• Envision language: a domain-specific language for supply chain constraints such as MOQs and BOMs.
• Supply Chain Scientist: a dedicated expert who implements and operates the solution.

• An electronics retail chain in Vietnam can automate store replenishment and DC allocation.
• An airline maintenance unit can size rotable and spare parts stock against aircraft-on-ground risk.
• An auto parts distributor can forecast demand per branch for a large catalog.

• Lokad reports over 20 billion USD worth of merchandise optimized on its platform.
• Probabilistic forecasts quantify the likelihood of every demand scenario for risk-adjusted decisions.
• Routine reorder, multi-echelon allocation and lead-time calculations are fully automated.
• Recommendations are tied to cost structures and service targets to cut financial waste.

Retailers, distributors and MRO operators face long-tail catalogs and erratic demand that classic forecasts handle poorly. Probabilistic, decision-centric optimization is gaining ground as companies in Southeast Asia expand store networks and spare parts operations.

Founded in Paris in 2008 by Joannes Vermorel (CEO) and Estelle Vermorel (CFO). Initially funded by love money and a few business angels, with no further investors; the company reports average growth of about 50% per year. Website available in English, French, German, Spanish, Italian, Japanese and Russian.