Minimal Code Systems AG is a Swiss-headquartered engineering company. Our core team has worked together since the 1990s — more than thirty years of computing, programming, networking and large-scale systems — designing and operating large systems and the infrastructure underneath them. We have been building and operating crypto exchange infrastructure since 2017. We run our own data-centre cage space and systems in Zurich, Frankfurt and Istanbul.
This page carries no client logos and no case studies. Keep scrolling — we explain why, and what we offer in their place.
Minimal Code Systems AG is a Swiss company, registered in Regensdorf in the canton of Zurich. You can check our legal details — company name, legal form, registered office and UID — before you read another word. We would rather you did.
Behind the entity is a small senior team of engineers who have been working together for a very long time. Thirty years ago that meant networks, operating systems and systems that had to stay up because there was no second one. Since 2017 it has meant crypto exchange infrastructure: matching, wallets, custody, liquidity, the nodes underneath and the operations around them. The subject matter changed. The discipline did not.
Other companies bring us the problems that require deep engineering — the ones that have already been escalated somewhere else. The engineers who take those problems on are the engineers who would be assigned to your account.
We build our own products, on our own stack, and we operate them ourselves. We also build bespoke systems for banks, investment intermediaries, trading firms and organisations with serious infrastructure to run, on the same foundations we already run in production.
| Swiss company | Minimal Code Systems AG · Regensdorf, canton of Zurich · UID CHE-206.432.155 |
|---|---|
| Core team | Working together since the 1990s · 30+ years in computing, networking and large systems |
| Exchange infrastructure | Built and operated since 2017 |
| Our own cage space | Zurich · Frankfurt · Istanbul |
Engineering culture is easy to claim and hard to describe. Here is ours in seven statements, each of which has a consequence you can test in a technical conversation.
Our trading, liquidity and blockchain products are our own code, written by our own engineers. Nothing central is licensed in and re-sold with a margin on top. The practical consequence is that when you ask for a change to how the matching engine behaves, how a KYC level is scored, or how a book is built, the answer is a conversation about design and priority — not a support ticket to a third party whose roadmap you do not control.
A licence and a handover document is not a system. We operate our own data-centre cage space and systems in Zurich, Frankfurt and Istanbul, with a 24/7 network operations centre behind them. The people who wrote the code and the people who are woken up by it work in the same company. That arrangement is uncomfortable in exactly the right way: an architecture that is clever on a diagram but painful at three in the morning does not survive here, because we are the ones awake.
Every dependency is somebody else's roadmap, somebody else's licence renewal and somebody else's failure you cannot fix. So the exchange platform ships with its own KYC, its own wallet and custody layer and its own admin, and it needs no third-party product to operate. If your risk team requires a named provider — Sumsub for identity, Fireblocks or BitGo for custody — we integrate one, and that stays a commercial decision rather than a technical constraint. Independent by default, integrated by choice.
The interesting question about a system is never how it behaves when everything works. Redundancy, failover, degradation behaviour and disaster recovery are designed at the start, across three locations, because they cannot be added credibly at the end. Migrations are staged and planned to avoid interruption, and validation happens without touching live systems.
We select, procure, install and operate the hardware our platforms run on, and we keep a hardware research laboratory for exactly that reason. Storage layout, network fabric and GPU topology decide what the software above them can promise. Teams that treat hardware as a purchasing exercise discover this late, usually under load.
We publish no throughput figures, no latency figures and no client logos. Instead the product is open in your browser and the load test is run with you, on scenarios you choose, before you commit to anything. It is a slower way to sell and a much faster way to find out whether we are right for you.
Not every workload belongs on private cloud, and not every buyer needs the platform we would most like to sell. If the honest conclusion is that you should stay where you are, or that another kind of supplier fits you better, that is what you will be told. It costs us a proposal and saves you a project.
We do a few things properly rather than ten things approximately.
Our own trading and blockchain products: a white-label crypto exchange with spot, futures and perpetuals and margin in one platform; a liquidity and market-making system that fills and hedges books across venues; and our own multi-chain node infrastructure behind a single API. We have been building and operating this class of system since 2017, and the same foundations carry bespoke work for banks, intermediaries and trading firms.
Blockchain & FinTech · White Label Crypto Exchange · MC Trading Infrastructure · MC Chain Infrastructure
Managed private cloud, dedicated GPU platforms and a secure DevOps environment, run from our own cage space in three locations. We design for your workload rather than to a price list, take on installation, monitoring, maintenance, security and capacity management end to end, and cut public cloud bills by up to 50% where the workload genuinely suits it. In these infrastructure services our team does not see your data, does not access application content and does not enter your services.
Cloud & Infrastructure · Managed HCI Private Cloud · Private AI GPU Cloud · DevOps-as-a-Service
SAP Signavio work — process mining, process modelling, process transformation and governance — delivered by consultants who have worked on large-scale engagements. Alongside it, an upcoming product that takes incoming invoices out of a shared inbox and puts them through a controlled approval workflow and into SAP.
Business Process & Automation · SAP Signavio Consulting · InvoicePipeline
One company with a Swiss home and an engineering and operations footprint spread across three cities.
Switzerland is where the company is registered and where its commercial and legal centre sits: Minimal Code Systems AG, Bahnstrasse 58, 8105 Regensdorf, in the canton of Zurich.
The engineering and the operations are deliberately distributed. We hold our own cage space and systems in Zurich, Frankfurt and Istanbul — three locations, three power and network environments, three jurisdictions of physical presence. That is what makes high availability, geographic redundancy and disaster recovery something we can design rather than something we hope for. It also means you can choose where a workload physically sits, which for a regulated operator is usually the first question rather than the last.
Around those locations sit our engineering teams, a hardware research laboratory and a 24/7 network operations centre. Between them they cover the whole stack we sell: matching engines and order book aggregation, blockchain nodes, hypervisor, storage, network fabric, cage and power, GPU clusters, CI/CD and SAP process design. We do not subcontract the difficult part and then manage it at arm's length.
| Location | What is there |
|---|---|
| Zurich | Registered office of the company, and our own data-centre cage space and systems |
| Frankfurt | Our own data-centre cage space and systems |
| Istanbul | Our own data-centre cage space and systems |
| Across all three | Engineering teams, a hardware research laboratory and a 24/7 network operations centre |
You expected a row of grey logos about now. A wall of client marks, three case studies with a percentage in the headline, a testimonial from a job title. There are none on this page, and there are none anywhere else on this site. That is a decision we made, not a gap we have not got round to filling.
There are two reasons, and both of them are in your interest as well as ours.
The first is discretion. Our work sits close to regulated, competitively sensitive operations. When a venue launches on our platform, the fact that it did not build the system itself is useful information to its competitors — and sometimes to its regulator's timetable. Discretion is part of what a client buys from us. We are not willing to spend it on decoration.
The second is that a logo proves nothing. Anyone can put a wordmark in a grid. You cannot verify it, you cannot phone it, and it tells you nothing about whether the system stayed up on a bad morning or whether the vendor answered when it did not. The same goes for a throughput figure produced on somebody else's hardware under somebody else's load pattern.
So we do not ask you to believe us. We would rather be checked. Four things are open to you before you sign anything.
The exchange platform is live and open. Not a recorded walkthrough, not a guided tour with a sales engineer driving: the actual product, in your browser, with spot, futures and perpetuals, margin, Earn, P2P, the account and security surfaces and the growth modules all present. Every capability we describe on this site exists in it. Go and try to break it.
We do not publish performance numbers. We prove performance on your terms. Before you commit, we run a live load and stress test of the platform with you: you define the scenarios that actually worry your risk committee, your engineers watch the real system under real load, and you keep the results. If it does not do what you need, you will have found that out before signing rather than after.
Request a Proof of Performance session
This is the part people assume is missing. It is not. We do not publish client names — and named references are available on request. Have a conversation with us, tell us what you are evaluating, and we will introduce you to clients whose situation resembles yours, under NDA. Not a curated quote in a box: a person at a comparable organisation, who can tell you what the delivery was like and what went wrong as well as what went right. What we will not do is publish their names on a website so that we can decorate it.
You will not be handed to an account manager reading from a briefing sheet. The people who answer your technical questions are the people who designed the systems and who will be reachable once yours is running. Bring your hardest question to the first conversation. It is the cheapest test either of us will run.
Bring us your hardest question
We would rather be checked than believed. Everything above is available before money changes hands — and any supplier worth its fee should be able to offer you the same four things.
And the things we will never put on this page: a performance number you cannot reproduce, a certification we have not been through, a partner tier we have not earned, or a client name we were not given permission to use.
We have not put names and photographs on this page yet, for the same reason we have not put logos on it: we would rather publish something when we can stand behind every word of it than fill a grid.
What we can tell you now is how it actually works. You deal with a small senior team. The engineer who answers your first technical question is the one who designs your system, and they stay on the account after it goes live — they are not a pre-sales resource who disappears at contract signature. Their background is thirty years of building and operating systems that were not allowed to fall over, and since 2017 that has meant exchange infrastructure specifically. Ask them about a design decision and you will get the reasoning, including the parts they would do differently now.
If you want to meet the team before you commit, say so and we will arrange it. That is a better introduction than a headshot.
Minimal Code Systems AG, a Swiss Aktiengesellschaft with its registered office at Bahnstrasse 58, 8105 Regensdorf, Switzerland, registered in the Commercial Register of the Canton of Zurich under UID CHE-206.432.155. The full details are on the legal notice, where you can check them against the register yourself. Contracts, invoices and liability sit with that entity.
Small and senior, by design — and a number would mislead you in either direction. A large figure says nothing about who is assigned to your work; a small one says nothing about how long they have been doing it. Three questions are more useful, and we answer all three in the first conversation: who will actually be on this account, what have they built before, and can I speak to them now. The relevant fact about scale is that other companies consult us on problems that require deep engineering, which is not a position you reach by having many people.
Because publishing them would spend a client's discretion for our benefit, and because a logo is not evidence. It is a deliberate policy, not an oversight — the section above sets it out in full. Named references are available on request: tell us what you are evaluating and we will introduce you to clients in a comparable position, under NDA.
Yes, and we would prefer it. The first conversation is technical, not a pitch. Come with the architecture question, the compliance question or the migration question that has been bothering you, and you will get a direct answer — including "that is not something we should do for you", when that is the honest answer. Talk to an engineer and we reply within 2 business days.
We make no compliance or certification claims on this website, on purpose — a vendor cannot hold a supervisory status on your behalf, and a badge in a footer has never protected anybody. What we can describe is how the environments are built: access controlled with authorisation, MFA, VPN, logging and security policies; isolated and auditable environments that take requirements such as PCI-DSS into account; and the operating rule that in our infrastructure services our team does not see customer data, does not access application content and does not enter customer services. Our responsibility there is uptime, performance, security, capacity, network and operational continuity. Where a customer separately asks us to operate an application as well — Managed Exchange Operations on our trading platform is the main case — that is a different service and we describe it differently: access is scoped to named individuals, authenticated, logged per action, reviewed on a schedule, and written into the agreement rather than assumed. Send us your security questionnaire and we will answer it in writing, question by question, before you have to make a decision.
In the environment and the locations you agree with us, from Zurich, Frankfurt and Istanbul. Location is a decision you make with us at design time, not something you discover afterwards, and geographic redundancy across those sites is what supports disaster recovery and business continuity. Access is controlled with authorisation, MFA, VPN and logging, granted to named individuals rather than shared accounts, and in our infrastructure services our team does not enter your services or read your application content. Where administrative access originates is a separate question from where the data sits, and we answer both — the access model is set out in full on the Cloud & Infrastructure page.
Open the platform and use it. Bring us the scenario that worries you and let us load-test it in front of you. Ask for references and we will make the introductions under NDA. If at the end of that you would rather work with somebody else, you will at least have made the decision on evidence.
We reply to enquiries within 2 business days, and the reply comes from someone who works on these systems.
