Technology In A Nutshell
Table of contents
They use our tech
A sound basis: Lokad's Platform
At the core of Lokad’s platform is Envision, Lokad’s tailored solution to the unique challenges of supply chain optimization. It provides the programmable power of a modern language with the usability of Excel, enabling business users and Supply Chain Scientists—not just developers—to build, deploy, and maintain sophisticated decision-making apps at scale, from dynamic pricing to inventory replenishment.
Lokad’s platform is designed to scale massively, managing data sets with hundreds of millions of SKUs or orders. It encourages smart automation aligned with business logic, instead of rigid or superficial automation. Additionally, it has been designed to offer a production-ready, secure, and upgradable environment and to reduce maintenance challenges.
The importance of data engineering
Lokad’s solution brings predictive optimization to supply chains by leveraging raw, unfiltered historical data and domain-specific programming. It emphasizes flexibility, automation, and practical implementation over rigid data standards or traditional BI pipelines.
Lokad ensures the integrity of client data through a rigorous two-tier health check process: low-level checks validate internal consistency within and across tables, while high-level checks confirm alignment with KPIs and business features critical to demand forecasting and optimization. The whole process is rigorously documented.
Forecasting and Differentiable Programming
Lokad uses probabilistic forecasting to model the full range of possible future scenarios (demand, lead times, returns, etc.), enabling more informed, risk-aware decisions. When combined with differentiable programming, forecasting models adapt and improve to new data flows so your supply chain decisions evolve with real-world conditions.
Differentiable Programming represents a paradigm shift in supply chain technology. Building on deep learning, it brings unmatched expressiveness, flexibility, and scalability and succeeds by embedding domain knowledge and aligning with specific business constraints, unlike general-purpose AI tools.
Envision: our secret code for financial optimization and automation
Envision is our domain-specific language (DSL) built for predictive supply chain optimization. With this, our supply chain scientists turn complex constraints and business goals into financially optimized and automated decisions.
Additionally, Envision is publicly available and documented. It targets users with basic programming knowledge. It emphasizes automated script rewrites, allowing seamless updates without manual changes. A key design principle of Envision is correctness by design, aiming to catch errors at compile time rather than runtime, which enhances productivity and system reliability, especially when handling large datasets.
Humans and machine:
your Supply Chain Scientist copilot
AI and full automation don't replace human support and cooperation. Our team of Supply Chain Scientists is there to implement your solution, maintain its effectiveness, and answer all your questions.