--- title: How Contember Fabrika works, in detail description: "Working with Contember Fabrika from the first call to everyday operations: the offer, the kickstart week day by day, what you own on Friday, and how it keeps living after." url: "https://www.contember.com/factory" type: static generatedAt: "2026-07-28T16:34:56.540Z" --- # How Fabrika works, *step by step* From the first call to everyday operations: what happens before the kickstart, what the week itself looks like, what you're holding on Friday, and how it all keeps living after. ~10 minutes of reading · prices are in the [pricing overview](/pricing) ![](/illustrations/strip-stations.webp) ## It starts with a half-hour call ### We go through the process that slows you down. You describe how it runs today, we ask who touches it and where it breaks. And we tell you right on the call whether it fits a kickstart. If it doesn't, you hear it right away, not after three meetings. ### Within a few days, you get an offer. It states the scope of the first application: what it will do on Friday and what it deliberately won't. It states the timeline. And it states the price. No estimates, no “we'll see how complex it gets”. Once you agree to it, the price doesn't move. ### You pick your person. Someone who knows your processes and isn't afraid of software: a capable power user, not a programmer. They'll be in the room from day one, because in time the next applications get built by them. Once you approve the offer, the kickstart usually starts within a few weeks, depending on calendars on both sides. And it's also a test of working together: within a week you know whether our way of working suits you, before you commit to anything long-term. ![](/illustrations/tall-callbox.webp) ## The kickstart week, day by day A week is enough because the scope is bounded up front: one workflow, one application. And because we don't start from zero: sign-in, permissions and deployment already stand, nobody invents them again. The week is the default, not a rule. A small app can be done in a day or two, a bigger one asks for more time. The offer shows how it comes out for your process. 1. Mon ### Process mapping We sit down with the people who work with the process every day, not just with management. We walk through how the workflow really runs, including the exceptions and workarounds nobody mentions in meetings. The output is a data model and a clearly agreed boundary: what the application will do on Friday, and what it deliberately won't. 1. Tue–Thu ### Building the application We build on our platform, with AI assistance and under the control of experienced developers. Every day we send you a link to the current state. The application grows in the open, not behind a curtain. Feedback goes in right away. Big changes to the brief belong in the next iterations, not in the middle of the week. 1. Fri ### Deployment to production The application is deployed into your infrastructure in a standard, repeatable way. We set up role-based access, train the people who will work with it, and hand over the documentation. From Friday on you work with the application for real, and from that moment we keep it running. What it asks of you: A few hours on Monday from the people who run the process every day. Your chosen person stays close to the build through the week, though not full-time. And on Friday, an hour or two for training. What the kickstart is not: A workshop, a feasibility study or a click-through prototype. At the end there is software your people work with. ![](/illustrations/tall-week-line.webp) ## What gets made, and how fast The pattern for anything where one person asks and another approves: holidays, business trips, small purchases. Estimate: A day or two The pattern for any register: contracts, vehicles, equipment, certificates. A record card, deadlines, reminders. Estimate: A day or two Field entries on a phone: trips, work done, quick reports. Anything written on paper today and retyped later. Estimate: A day or two A document arrives, makes its approval rounds, gets matched. A purchase request or a production requisition works the same way. Estimate: Two or three days Enquiries, offers, open deals. Not a hundred-field monster: exactly the steps your sales actually takes. Estimate: Two or three days ![](/illustrations/tall-app-vending.webp) Hours by project, straight into billing. The same shape works for machine logs, shifts or service visits. Estimate: Two or three days A login for your clients: order status, documents, requests. Your app facing out, not just in. Estimate: Four or five days An order from intake to delivery, with statuses, history and follow-up steps. Whether you sell goods, services or production. Estimate: Four or five days People, machines or dates against commitments. From production shifts to consultant workloads. Estimate: Four or five days Anything that lives in spreadsheets, e-mail and people's heads today. During the kickstart, a small app is done in a day or two, a bigger one takes the whole week. Nine examples out of many. It works the same in manufacturing, services, commerce or an IT company. The app follows your process, not the other way around. [Pick your first app with us →](https://cal.com/contember/demo) - helpdesk - bookings - complaints - inspections - onboarding - board reporting - shift handovers - asset register - purchase requests - training records - stock counts - visitor log - helpdesk - bookings - complaints - inspections - onboarding - board reporting - shift handovers - asset register - purchase requests - training records - stock counts - visitor log The list has no end. The scoping call is where we pick what makes sense for you first. [Book a 30-min call →](https://cal.com/contember/demo) ## On Friday you're holding more than an application On Friday you take over an application in production: access set up, people trained, documentation handed over. But the more important part of what you now own sits underneath. Let's go through it layer by layer, top to bottom. Your application what your people see Platform the production line Access management roles & permissions Operations deploys · backups · updates Your infrastructure your cloud account ![](/illustrations/sym-app.webp) ### Your application Forms, overviews, roles, a history of changes. This is what your people start working with on Friday. It looks like your process, because it was built from it. And it's the first product of your fabrika, not the last. ![](/illustrations/sym-engine.webp) ### Platform The fabrika's production line: a tested foundation and established practices we've built from years of making business software. It's why the first application stands within a week, and why every next one takes days. Nothing that already stands gets invented again. ![](/illustrations/sym-turnstile.webp) ### Access management Who may do what, in one place for all your applications. Someone joins, leaves or changes roles: one change, not five tools. And every action leaves a record. ![](/illustrations/sym-gauges.webp) ### Operations Deployment, backups, updates, monitoring. The application keeps running long after the initial project enthusiasm fades. That's our job and a contractual commitment, not an extra task for your people. ![](/illustrations/sym-tower.webp) ### Your infrastructure All of it runs in your own cloud account, in your company's name. The code and the data belong to you, and you can leave us at any time, with the whole system. [How it works technically: the Under the hood chapter →](#tech) These aren't tools bought in for one project. We build our own products on them too, like Nua Site and Redo My Site. That's why we think like manufacturers, not like an agency. [What else we build →](/products) ## After the kickstart, the next applications get built in-house Fabrika isn't outsourcing where everything disappears behind a vendor's wall. The work is divided so that competence gradually moves over to you. This is how it usually goes: 1. wk 1–2 ### Finding their feet We keep operations running: deployment, access, monitoring, platform updates. Your person saw during the kickstart how the platform is used; now they get their hands on it. The first small things we solve together. 1. mo 1 ### First changes of their own After a few weeks your person makes smaller changes themselves: fields, code lists, reports, roles. We review every change they send toward production. The time investment is a few hours a week, not a new full-time job. 1. mo 3+ ### Their own applications In time they build whole applications themselves, always with our review behind them. They grow with it: they build things the entire company uses every day, and that competence stays with you. Your dependency on us shrinks over time instead of growing. 1. ongoing ### Us behind your team, leadership in the picture We stay where we're most useful: supporting your team and contractually guaranteeing operations. There's always someone to call. Leadership gets a monthly overview: what's running, what changed, what's in the backlog. And the costs are known up front: the kickstart price from its offer, then a predictable monthly plan. The plan gets agreed only after the kickstart, once you know you want to continue. ![](/illustrations/tall-growing-rack.webp) ## Under the hood This chapter is for whoever you forward this page to with the words “have a look at this”. Everyone else can skip ahead, the comparison will wait. ### Where it all runs In your own Cloudflare account, in your company's name. The applications run as Cloudflare Workers, data lives in a D1 database (SQLite), files in R2 storage. We act in the account as administrators, and you can revoke that access at any time. ### What it's built from No no-code: every application is ordinary code. TypeScript and React on standard web technologies, plus our own open tools for building, sign-in and deployment, developed as open source. We build with AI assistance, inside guardrails and under the review of experienced developers. The code lives in git and every change gets checked before it ships. ### One sign-in for everything People sign in with their company account over SSO (OIDC, so Google, Microsoft and others work). Roles and permissions are managed in one place across all applications, and every action leaves an audit record. Machines and connected systems get API keys instead. Internally we call this layer Propustka. ### Deployment without clicking around Every application describes what it needs in a single file: domains, databases, permissions, build pipeline. Deploying to production is a repeatable step, not clicking through a console. On top of that sit monitoring, updates and backups, and the database can be restored to any point in the last 30 days. Internally this layer goes by Vozka. ### What the servers cost You pay Cloudflare directly, by actual usage. Internal applications barely register: typically a few dollars a month for the whole fabrika, tens at most. And nobody charges per user, Fabrika plans scale with the number of applications, not seats. ### Integrations and existing data Connecting your accounting, ERP or e-shop is technically not a problem, and neither is taking over the data you have today. What differs is the effort: spreadsheets migrate quickly and simpler integrations fit inside the kickstart week, bigger systems ask for further iterations. How it looks in your case, we can tell you up front on the first call. ### If it ever comes to leaving The application code is yours, the data and domains sit in your account, and the platform stands on open tools. The system keeps running without us, and anyone who does ordinary web development can take it over. Nameplate Runtime Cloudflare Workers Database SQLite (D1) Files R2 storage Code TypeScript + React Sign-in SSO (OIDC) Deploys git + review Platform open source Your account, your code, your data. ![](/illustrations/tall-nameplate.webp) ## What you're hiring instead of us today We don't compete with brands, but with the ways this usually gets solved today: staying in spreadsheets, buying an off-the-shelf tool, clicking apps together in a no-code platform, or having a system custom-built. Each of them makes sense somewhere, and somewhere it stops being enough. The real question isn't what's cheapest now, but what runs your company three years from now. ![](/illustrations/machine-carts.webp) | What we compare | Stay with spreadsheets and email | Buy an off-the-shelf tool (SaaS) | Click it together in a no-code platform | Hire a custom-build agency | Contember Fabrika | | --- | --- | --- | --- | --- | --- | | First version in production | It's already there, and that's the problem | Weeks, but on someone else's template | Days, but only what the boxes allow | 3–6 months | 1 week | | Cost and speed of change | Free, but with no control or history | You wait for the vendor's roadmap | Yourself and right away, within the tool's limits | Tens of thousands, months of waiting | Included in the monthly plan, ongoing | | Ownership of code and data | Nobody really knows | The vendor | There is no code; data sits on their platform | Often stays with the agency | You: code, data, portable | | Access management and audit | Shared files and passwords | Separately in every tool | Basic, whatever the platform offers | It varies | Central, with every action audited | | Continuity | Ends with the person holding it together | Ends with the product or the price | Ends at the first process that doesn't fit the boxes | Ends with the contract | Your own trained person and a clear way out | ### “I've seen AI coding. But who's going to maintain it here?” These days almost anyone can generate something: a nephew, a freelancer, a capable colleague. And it often even works. But without access management, proper deployment and maintenance, it becomes shadow IT that ends the day its author leaves. AI made writing software cheaper. It didn't make owning it cheaper. That second half is exactly what we do. [Discuss it on a 30-minute call →](https://cal.com/contember/demo) ## Why this isn't another dependency We hear these questions at every first meeting, and they're fair. ![](/illustrations/spot-exit.webp) ### “We don't have IT people. Who here will know how to run this?” That's what the whole model is built around: we build the first application and your person is there with us. They build the next ones themselves, with our review and support behind them. A capable power user who knows your processes is enough. You don't need a programmer. ### “What if you shut down or raise your prices?” The code and the data belong to you, we build on standard technologies and everything runs in your infrastructure. You can leave us at any time, with the whole system. It sounds odd as a sales argument, but it's exactly why clients stay with us. ### “Custom development means half a year and millions.” Traditionally, yes. That's why we bounded both the scope and the price up front. One week, one workflow, one price you know before we start. The speed isn't magic. It comes from not starting from scratch: the platform is already built, on years of experience building business software. ### “Are these just web apps for the office?” The apps run in a browser on anything: on a computer in the office, on a phone or tablet in the field. A trip log gets filled in in the car, a handover report on site. And when a full mobile app makes sense, that can be built too. ### “Can it connect to our accounting, ERP or e-shop?” Technically that's not a problem, and neither is taking over your existing data, spreadsheets migrate quickly. What varies is the effort: simpler connections fit inside the kickstart week, bigger systems ask for further iterations. How it looks in your case, we can tell you on the first call. ### “We don't have a cloud account. Now what?” You don't need anything up front. We set the account up, in your company's name, and hand it over as yours, along with everything that runs in it. Managing it stays part of our job. ### “Where does our data flow when AI is assisting?” The application runs in your infrastructure and AI always runs through your own plan, your people's AI, so you know exactly where things flow. We don't train anything on your data. And we're happy to advise on which provider makes the most sense for you. ### “How many users and how much data will it take?” More than an internal application will ever need. It runs on Cloudflare's global network, the same infrastructure that serves a large part of the internet. When the company grows, the applications grow with it, no replatforming. ### “Won't this just be another vendor dependency?” We won't promise you complete independence. Nobody honest will. But the model points the opposite way from a classic vendor: the capability to build and manage the apps grows on your side, we support your team long term and hold an SLA on operations. Every action leaves a record, everything is documented, and you can take the whole thing with you. The way out is clear. ### “What if it doesn't catch on internally?” That's why the entry step is a Kickstart, not a year-long contract. Within a week you see a working application on your own process, with your own people. Only then do you decide whether to continue. ![](/illustrations/cta-crane.webp) ## Your first app can be live in a *week*. The best first step is a half-hour call. You pick the process that slows you down the most today, and we tell you whether it fits the kickstart, or say plainly that we're not the right match. [Book a 30-min call →](https://cal.com/contember/demo) [Or just write to us →](/contact) Capacity: 2–3 kickstarts a month. Slots fill up gradually.