We are a metals company that happens to write software.
KNet Technologies B.V. is based in Amstelveen, the Netherlands. We build manufacturing execution, scheduling and planning software for steel, aluminium and copper producers — and we came to it from the plant side, not the other way round.
The gap in the middle of the market
There are excellent metals platforms in this industry. There are also plenty of mid-size producers who will never be able to buy one — not because the software is not worth it, but because the licence, the database underneath it and the implementation programme together add up to a number no mid-market board will sign.
Those plants do not stop needing execution software. They run on a patchwork of spreadsheets, an L2 system doing three jobs it was never designed for, and one person who knows how it all fits together. It works until that person retires.
KNet exists for that gap. Same functional depth at the execution layer, an open-source data stack that removes the licence line entirely, and a deployment model a plant of 400 people can actually operate.
Background
Our domain experience was earned across major integrated and mini-mill producers — including Tata Steel Netherlands, JSW Steel, SAIL and Essar Steel — in operations, MES and Level 3 programme roles.
Five things we hold to
Defer to the plant
When our model and an operator's experience disagree, the operator is usually right and our model is usually missing a constraint. We change the model.
Phased, reversible rollouts
One unit or one module first. Prove it in production, then extend. Every phase should leave you better off even if you stop there.
Boring technology where it counts
PostgreSQL and a modular monolith are not exciting choices. They are choices you can still operate in year eight with a small team.
Say when we are not the answer
If your requirement is really an ERP, or you need fifty reference plants today, we will tell you at the first meeting rather than the fifth.
No lock-in theatre
Source code escrow, open data formats, documented interfaces and per-customer deployment. If you ever want to leave, you should be able to — that constraint keeps us honest about earning the renewal.
Working with a small team
We are deliberately small and deliberately specialised. That has consequences worth being upfront about:
- You talk to the people who write the software, not an account layer
- Decisions get made in a day, not in a change control board
- We cannot run six parallel implementations — and we will say so
- Source code escrow exists precisely because we are small
Domain people, plus the engineers to build it
The core team combines steel operations and MES functional experience with software engineering across Django, PostgreSQL, React and Azure infrastructure. We extend it with domain-screened specialists — including PSI Metals consultants — for client engagements.
Building a metals software company, not a project shop
Consultancy and resourcing fund the work. The product is the point.
Product
DigiPlantMES, with its scheduling and planning modules, developed against real plant requirements through the Design Partner Program.
Services
Consultancy, transformation and development work that keeps us in plants — which is where the product requirements come from.
People
A network of metals domain specialists we place onto client programmes, and who feed real problems back to us.
Come and argue with our assumptions.
The most useful conversation we can have is with someone who will tell us where the model is wrong. If that is you, the design partner programme was built for it.