Selected work
A selection of engagements in supply chain, production, industrial systems, administration and AI adoption. For each project we highlight the problem, the change introduced and what changes in the work of those who run the process.
Supply chain / Decision Intelligence
Demand planning in a manufacturing context where uncertainty directly affects inventory, stock-out risk, tied-up capital and supply reliability.
Exceptions to review
01 · Context
A single-point forecast did not show when demand was stable enough to support a purchasing or reorder decision.
02 · Intervention
Forecasts with a measure of uncertainty, reliability assessment, safety stock and reorder point calculation, with priority on the cases that need attention.
03 · Operational effect
Planners see forecast, reliability, risk and an operational proposal, so they can focus their time on exceptions and the most relevant decisions.
The company needed to use generative models and internal knowledge without exposing data or losing control over usage. We designed an internal infrastructure that governs access, authorisations, sources used, traceability and monitoring, keeping data and accountability within the organisation.
Process control required linking what is observed on the product to what happens during machining. High-resolution vision, laser reproduction, process data and simulation were integrated into a single system supporting quality, analysis and development.
In document processes a large share of time went into reading, entering and checking information that was already available. The system captures documents and requests, extracts the data, applies known rules and prepares the next step; people step in on validations, exceptions and decisions.
Other systems delivered
Whether in production, quality, sales or administration, we start from a performance, a process or a decision that must improve. Technology is chosen only after understanding what produces the result and who is responsible for it.
We connected KPIs, progress, traceability and line utilisation to the way production is actually planned and managed, making available capacity, deviations and operational priorities more visible.
To reduce the time needed to reconstruct cases and evidence, AI retrieves useful knowledge and supports analysis following FMEA and 8D methods. The expert retains technical judgement, approval and final responsibility.
Data and documents are captured automatically, known rules are applied and the system prepares the order or quote. The responsible person focuses on incomplete, inconsistent or out-of-policy cases instead of checking every file by hand.
We automated the capture, verification and processing of information in administrative and accounting processes, reducing repetitive entry and checks while keeping the evidence and traceability needed for control, audit and review.
Your problem will be different
We are most interested in problems where the result depends on the interaction of several functions, decisions, data and systems, and a standard solution cannot explain why performance isn’t improving.
Let’s discuss your context
