ERP · AI · Supply Chain · Quantum Computing — insights from building a real enterprise learning lab
Add, edit, bulk-load and audit real rows in DBX4 — a database engine written from scratch in C/C++ — now live on the portal.
created_by and updated_at answer who and when. They do not answer "which session" — here is the fifth column that does.
DBX4 is a relational engine written from scratch in C++ — no dependencies — where every row records who changed it and in which session, events drive the accounting workflow rather than logging it, an
How to let an AI assistant work with ERP data responsibly: read-only by design, human-in-the-loop writes, self-explaining responses, and defence in depth — the principles that separate a useful AI int
Four forecasting methods backtested per item with MAPE, a 3-month projection with error band, and a safety-stock formula wired to real lead times — the demand signal MRP consumes.
The portal's first write path into BOM data: edit structures with live cost roll-up, automatic version bumps, and guardrails that reject self-parents, circular structures, and duplicate lines.
Running Material Requirements Planning on live data and watching the plan change — planned orders, exceptions, and coverage recomputed across an eight-week horizon.
Following the work orders the S&OP plan created: the WO list, one order opened with its BOM materials and routing auto-loaded, and the release-to-complete lifecycle — with live screenshots.
Reading BOM-001 for the SmartBar Pro 500: the 4-component structure, Level 2 raw materials, on-hand vs available, and the live cost explosion — with live screenshots.
RTG-001 in Flow View and Full Detail: setup vs run time, compounding yields, departments and cost rates, and the exact match with WO-2025-0001's planned hours — with live screenshots.