Frequently Asked Questions

We utilize a structured 6-month delivery roadmap designed to validate value early and ensure a secure, user-validated rollout:

Phase 1 - Proof of Value (The Pilot) — 3 Months We focus on "ROI First." To bypass immediate IT hurdles, we build a custom optimization model using your actual static data. The goal is not just to test the logic, but to validate the business case and identify specific operational or financial improvements before any major infrastructure commitment.

Phase 2 - Production Implementation — 3 Months We move to "Operational Readiness." This phase covers fine-tuning the optimization model, developing automated data integration (via ERP connectors or pipelines), performance tuning, and provisioning your dedicated containerized AWS production environment. It concludes with a dedicated User Acceptance Testing (UAT) period to ensure your team is trained and the system is ready for Go-Live.

WonForge uses a three-phase model: a fixed-scope Pilot to validate ROI on your real data, a fixed-scope Implementation to move to production, and a month-to-month subscription once live. There are no annual contracts and no auto-renew traps - we earn renewals through continuous results.

All models and data are hosted in secure AWS environments. Each client's data is isolated in its own S3 bucket, with encryption applied both at rest and in transit. Access is tightly controlled, and we follow AWS best practices for cloud security. We are also open to NDA arrangements tailored to your requirements.

We use structured data such as production capacities, inventory levels, bills of materials, and routing.
Most clients choose one of two integration options:
• Direct Connector (minimal client effort):
We build custom connectors or adapters to your ERP, MES, or database systems. This approach requires little to no ongoing effort on your side, since data flows automatically into the model.
• File/API Export (more client control):
Alternatively, clients can provide structured exports (CSV or API feeds) from their systems. This gives you more control over the data extraction process while still ensuring smooth integration into the model.

We focus on manufacturing and complex supply chains, particularly in specialty chemicals, performance materials (lubricants, adhesives, coatings), food & beverage, and other process industries. We solve demand, production, inventory, transportation, and raw material decisions together in one integrated, end-to-end supply chain.

Our models leverage an industry-leading commercial solver hosted on AWS, capable of handling millions of constraints and very large datasets. They are engineered for speed and reliability, delivering insights that drive both financial and operational performance.

While our core expertise is in optimization, we also develop complementary models to support broader decision-making. These include:
• Forecasting models - demand forecasting using statistical and machine learning approaches.
• Scheduling models - detailed, time-grained production schedules (minute-level Gantt charts) for execution and capacity planning.
• Rule-based or heuristic models - when optimization is overkill or data is incomplete, we can implement practical heuristic or rule-based approaches for faster deployment.

Standard ERPs treat equipment changeovers as fixed, average durations regardless of the production sequence. They cannot comprehend that cleaning a reactor after a dark, highly viscous batch takes eight hours, while following a clear batch takes one. WonForge uses mathematical optimization to sequence campaigns based on these exact physical realities, minimizing flush time and maximizing your capacity.

Most Advanced Planning and Scheduling (APS) tools use "heuristics"—simple, linear rules of thumb that schedule one order at a time. In changeover-heavy environments, this approach quickly creates bottlenecks. Mathematical optimization evaluates millions of variables simultaneously across your entire network, calculating the absolute best end-to-end plan that respects all physical constraints and maximizes your P&L.

In process manufacturing, if your holding tank is full, your reactor has to stop. You cannot plan them in isolation. Our optimization engine synchronizes your reactor campaigns directly with tank farm capacity, curing times, and outbound shipping schedules. This unified model ensures material flows continuously without bottlenecking your network.

Spreadsheets process information linearly and rely entirely on human intuition to spot conflicts. When a single raw material is delayed, a spreadsheet cannot automatically recalculate the optimal ripple effect across dozens of shared reactors and intermediate blends. A custom optimization engine does the deep computational heavy lifting, methodically evaluating millions of routing permutations to mathematically guarantee the most profitable revised sequence.

By solving the exact physical bottlenecks that drain profitability. Our models systematically reduce sequence-dependent changeover times, minimize off-spec transition waste, and optimally allocate constrained raw materials to the highest-margin orders. It stops revenue leakage on the plant floor while right-sizing your inventory.