Custom information systems development

We build company information systems for internal agenda — from process and data analysis through architecture design and ERP integration to operation and support. Systems your employees work in; applications aimed at customers and partners in a browser have their own section.

I am interested in a system
development-process
Selected references:
Hyundai Motor Manufacturing Czech
UNYP
Karlovy Lázně
Metrostav a.s.
More referencesshow more

Which page is right for you?

What a company information system is and when a boxed product falls short

A company information system is software your employees work in — it records the agenda, enforces approval workflows, keeps data in one place and produces the basis for decisions. Unlike an application for customers, nobody sees it from outside, and yet the daily running of the company rests on it.

A boxed ERP handles accounting, stock and payroll. Where it runs out is the specific agenda that makes up your competitive advantage — capacity planning by your rules, an approval process with five roles, a calculation nobody else does the same way. Those are exactly the things that end up in spreadsheets next to the system you pay licences for.

So the answer is often not replacing the ERP but extending it through an API, which can be significantly cheaper and faster than replacing the whole system. Applications aimed at customers and partners in a browser are a separate discipline; the boundary follows who signs in.

Comparison of a boxed enterprise system and a custom information system across six criteria
CriterionBoxed systemCustom system
AgendaStandard processesYour processes, exceptions included
Spreadsheets alongsideThey stayThey go, the agenda is inside
Roles and permissionsPredefined levelsDown to records and fields
IntegrationsWhatever connectors shipAnything with an interface
FeesPer user, per yearOne-off build, then operation
Changing a processYou wait for a vendor releaseShipped in the next iteration

The internal systems we build most often

Most engagements fall into one of these types, or combine two of them. Each gets its own page, including what comes up most often in that area.

CRM systems

Customer relationship management shaped to your sales process — opportunities, field rep activity, quotation and reporting on current data.

Custom CRM systems

Intranets and internal portals

One place for documents, policies, requests and internal communication. With approval workflows, versioning and access governed by the org structure.

Intranet development

Databases and data architecture

Data model design, migration of large volumes and performance tuning where the existing database can no longer keep up or data has fragmented into inconsistent copies.

Database systems

Logistics and warehouse systems

Warehouse management, shipment tracking, route planning and connections to scanners, GPS or sensors. For operations where late information costs money directly.

Logistics systems

Systems for organising people

Shift planning, attendance and HR management in large teams and production plants — with statutory limits and the rules every company has differently.

People organisation

Job and project administration

Digital workflow from enquiry to invoice, tracking cost, capacity and margin in real time instead of a month-end close glued together from several sources.

Job and project administration

Applications that people outside the company sign in to — client and supplier portals, reservation and ordering systems — belong to web application development. If the agenda also has to reach people in the field, we add a mobile application on the same backend.

A complete approach to developing large custom information systems

Digitising internal agenda with measurable returns

Digitising internal agenda with measurable returns

A custom company information system has one job: to save your people work they currently do by hand. Retyping between spreadsheets and e-mail, chasing who approved what, a month-end close glued together from five sources. We design the system around those specific spots, because that is where the return on investment shows up first.
Integration with ERP, CRM and third-party systems

Integration with ERP, CRM and third-party systems

An information system rarely stands alone. Through APIs we connect it to SAP, Helios, Microsoft Dynamics, Salesforce, accounting packages, banking interfaces and public registers. The goal is data flowing automatically and without duplicates — and nobody copying numbers from one screen to another. On selected projects we can safely expose an internal system to AI tools through the Model Context Protocol (MCP).
Process analysis, data model and data migration

Process analysis, data model and data migration

Before development we run a process and data analysis — mapping how the agenda actually works, not how a policy document describes it. From that come the data model and the architecture proposal. Migrating historical data from an existing system or from spreadsheets is a chapter of its own: cleaning, deduplication and verifying that balances and history still add up afterwards. Database design and data architecture have their own section.
Roles, permissions and internal data protection

Roles, permissions and internal data protection

In an in-house system, who sees what is critical. We design a detailed model of roles and permissions down to individual records and fields, backed by an audit log showing who changed what and when, and by multi-factor authentication at sign-in. We take GDPR requirements into account in the design, including access control, protection of personal data and retention periods set to what the particular project needs. For clients in finance and industry we pass their own security audits.
User training and an interface people do not avoid

User training and an interface people do not avoid

An internal system stands or falls on whether employees actually start using it. So we design the interface for minimal cognitive load — the most frequent task in the fewest clicks, understandable error states, no screens that need retraining every time. Launch includes training key users and handing over documentation. Our general approach to UX and UI is described in its own section.

How information system development works

Compared with an application for customers, internal systems have two extra phases — the ones projects most often fail on: migrating historical data and training the people who will have to work in the system every day. We plan both from the start, not with whatever budget is left.

  1. Process and data analysis

    Workshops with the people who actually do the work — mapping how the process runs, not how a policy describes it. The gap between the two is usually the most valuable thing the analysis produces.

    Phase output: a functional specification and a map of data flows.

  2. Data model and architecture

    Design of the data model, roles and permissions, and interfaces to surrounding systems. In an internal system the data model is a ten-year decision, so it gets its time before any code is written.

    Phase output: the data model, the permission model and an integration design.

  3. A prototype to click through

    A clickable prototype of the key screens, reviewed by the future users — we design it in Figma. Comments at this stage cost hours; the same comments after launch cost weeks.

    Phase output: an approved prototype and an updated specification.

  4. Development and integration

    Development in iterations with ERP, accounting and other systems connected as we go. We tackle integrations early, because that is where the surprises that move a schedule appear.

    Phase output: working modules on a test environment after every iteration.

  5. Data migration

    Cleaning and transferring historical data from the existing system or spreadsheets. First a trial migration into a test environment, where you verify on real data that balances, history and relations hold.

    Phase output: an approved trial transfer and a prepared rollback path.

  6. Training and go-live

    Role-based training, documentation handed over, and the live transfer usually over a weekend. In the first weeks we watch where users make mistakes and adjust the interface at those points.

    Phase output: the system live and the key users trained.

  7. Operation and growth

    After launch a service agreement takes over, along with development driven by priorities that emerge from real use. Large systems add project management.

    Phase output: a guaranteed response time and a development plan for the period ahead.

How much an information system costs

The ranges come from projects actually delivered and from the same model our price calculator runs on. By far the most common and cheapest option is the first one — an extension on top of what you already have.

Extension of an existing system

3–5 months of development

CZK 650,000 – 1,500,000

One specific agenda built alongside your ERP and connected through an API — approvals, capacity planning, a portal for branches. You leave the core that works alone and solve only what is missing.

A standalone company system

6–12 months of development

CZK 1,500,000 – 4,500,000

A system covering the whole agenda of one department — several modules, roles and permissions, reporting and integrations with surrounding systems. The most common scope of standalone engagements.

A company-wide system

from 18 months of development

from CZK 6,000,000

A system for the whole company across departments, with dozens of roles and connections to critical systems. At this size a range stops being useful; we price it after the analysis, usually in stages.

What drives the price of an information system most

  • The number of modules and agendas the system must cover — every additional area means its own screens, rules and reports.
  • The number of user roles and the granularity of permissions; a model down to field level costs far more than three access tiers.
  • Connections to ERP, accounting or public registers — days for a documented API, weeks for older systems with no interface.
  • The volume and quality of historical data; cleaning duplicates and inconsistent code lists usually costs more than the transfer itself.
  • The complexity of approval processes and calculations — multi-stage workflows with exceptions raise the price more than anything else.
  • The scale of training and go-live; running across several sites at once needs its own deployment plan.

The amounts are exclusive of VAT and cover analysis, design, development, testing, data migration and training. These are indicative ranges; the actual price depends on the specific brief, the extent of integrations, the quality of the data and other parameters of the project. The development times apply to a standard project scope — the real schedule is driven mainly by the range of features, the integrations, the availability of source materials and your team's input. For an indicative figure for your own brief use the software development price calculator; the exact one comes out of the process analysis. The contractual framework and code ownership are covered under custom software development.

Calculate an indicative price

Integration and migration: the riskiest part of a project

Given a well-defined brief, the development itself can be estimated relatively accurately. The biggest uncertainty usually sits in the integrations and the historical data — other people's systems behave differently from their documentation, and historical data appears in no documentation at all. So we tackle both as early as possible.

The systems we connect to

Twenty years have taught us a fair amount about how each platform behaves in practice. Most often these:

  • ERP and enterprise systems — SAP, Helios, Microsoft Dynamics, ABRA, Money.
  • CRM platforms — Salesforce, Microsoft Dynamics CRM and custom-built systems.
  • Accounting and invoicing packages, including document exchange in the ISDOC format.
  • Public registers and interfaces — data boxes, ARES, the commercial register, the VAT payer register.
  • Banking APIs for payment matching and bulk payment orders.
  • Carriers, warehouse systems, barcode scanners and GPS units.

Migration and go-live with minimal impact on operations

The live transfer is a one-shot operation that cannot be retried. We plan the migration and go-live so that the impact on your operations is as small as possible:

  • An audit of existing data — duplicates, missing relations, historically inconsistent code lists.
  • A trial migration into a test environment where you verify balances and history on real data.
  • The live transfer only after written sign-off of the trial run, usually over a weekend.
  • A prepared path back to the old system in case something only shows up in production.
  • For critical agendas, both systems running in parallel for an agreed period.
  • Role-based training and heightened support in the first weeks, when most questions arise.

Information systems that daily operations rest on

Four projects where our clients' employees use the system every day. The full list is in the references section.

System for client and network infrastructure management

For an internet service provider we built a system managing the client database, connection points, tariffs and payments. It records the technical infrastructure and connects to invoicing, so the commercial and technical agenda run in one environment.

Result: sales, technicians and invoicing work on the same data instead of three separate records.

Eurosignal case study

In-house system for an eye clinic

For the Horní Počernice eye clinic we are building an in-house web application for staff and operational agenda — employee document management, equipment records, planning of internal processes and overviews of HR and operational data. It includes a model of roles and permissions based on who works with which agenda.

Result: management and staff have the employee and operational agenda in one place.

IRIS case study

CRM for field sales representatives

A CRM system for a pharmaceutical company organising the work of field representatives — visit planning, meeting records and results reporting. It had to work for people who spend most of their time out of the office.

Result: management sees the sales network's activity as it happens, not in a monthly report.

Green-Swan case study

Point-of-sale system and gallery loyalty programme

For Galerie Rudolfinum we delivered a point-of-sale system joined to the loyalty programme's client management. Staff at the desk see a visitor's history and can offer individual terms; marketing builds its communication on the same data.

Result: the box office and marketing work on one client database.

Rudolfinum case study

Frequently asked questions about company information systems

We map existing processes, data flows and the needs of each role — usually in workshops with the people who actually do the work. The output is a functional specification, a data model and an interactive prototype you click through before we write a line of code. For large briefs the process analysis is a separate paid phase, and its output can also be used to approach another supplier.
Yes, integrations are our domain. We have experience connecting to SAP, Helios, Microsoft Dynamics, Salesforce, accounting packages, banking APIs and public registers. We design integrations with outages and retries in mind, to minimise the risk of data being lost or processed twice.
Migration is a separate project phase with its own budget. First we analyse and clean the data — duplicates, missing relations, historically inconsistent code lists. Then we run a trial migration into a test environment where you verify on real data that balances, history and relations hold. The live transfer happens only after the trial run is signed off, usually over a weekend, with a rollback path ready.
We design the model of roles and permissions down to individual records and fields, so a payroll accountant sees different data from an operations manager. Every change is captured in an audit log, and sign-in is protected by multi-factor authentication. GDPR requirements are taken into account from the design onwards — from access control through protection of personal data to retention periods set per project. For clients in regulated industries we pass their own security audits.
It depends on where the problem is. If the ERP handles accounting and stock but you are missing a specific agenda on top of it — approvals, capacity planning, a portal for branches — an extension connected through an API can be significantly cheaper and faster than replacing the whole system. A complete custom system makes sense where the existing solution is holding back the core of the business or nobody maintains it any more. That decision comes after the analysis, not before.
We involve key users in the analysis and the prototype, so the system is not a surprise but something they helped shape. Before launch we run role-based training and hand over documentation. After deployment we watch where users most often make mistakes or get stuck and adjust the interface at those points — adoption is settled in the first weeks of operation, not in a training room.
We hand over the source code, and for the parts of the solution we created you receive a perpetual licence to use them; open source and other third-party components remain subject to their own licence terms. That way you are not tied to a single supplier. The whole contractual framework and everything included in the handover is covered on the custom software development page.
The price follows the number of modules and user roles, the complexity of approval processes and the number of integrations with ERP, CRM or public registers. An extension on top of an existing system is significantly cheaper than an enterprise system built from scratch. Our software development price calculator gives you an indicative figure; the exact one comes out of the process analysis. After launch a service level agreement (SLA) with a guaranteed response time takes over.
An extension on top of an existing system is usually done in three to five months, a standalone system for one department in six to twelve, and a company-wide solution takes eighteen months or more. The analysis and data model take the first month or two. Large systems are deployed in stages — the first department goes live while the next is still being built, so you do not wait until the end for the benefit.
Yes, but it has to be accounted for in the data model from the start — retrofitting multi-currency or language variants into a finished system costs many times more than allowing for them at the outset. We handle data separation by site, separate numbering series and per-country rules, plus interface localisation. Which is why the analysis also asks where the company plans to expand.
Either works. We recommend the cloud for availability, backups and easy scaling, but for clients with regulatory requirements or their own server room we deploy on-premise too, or in a hybrid mode where sensitive data stays with you. The application runs in containers, so the choice of environment does not change how it is developed or deployed.
Basic reporting and overviews are built into the system so users do not have to export data anywhere. Beyond that we prepare a data interface for tools such as Power BI or Tableau, or replication into a separate analytics database so heavy queries do not load the operational system. Data architecture design has its own section.

Technologies we work with

We build enterprise systems on technologies that handle both heavy load and a long life. Backend on Ruby on Rails, frontend in React, data in PostgreSQL and integrations through documented APIs. We use widely adopted technologies with a long-established community and ecosystem, so the system connects safely to its surroundings and grows with the company. Our full technology stack is written up on its own page.

railsreactgraphqlnodejsaws
More about web and mobile application development technologiesshow more

Clients who trust us

Hyundai Motor Manufacturing Czech
UNYP
HiPP Czech s.r.o.
Magistra, a.s.
Webglobe, s.r.o.
Metrostav a.s.
Družstvo EUROSIGNAL
SEPOS, spol. s.r.o.
Galerie hlavního města Prahy
Galerie Rudolfinum
Brown-Forman Czechia, s.r.o.
ASSA ABLOY Entrance Systems, spol. s.r.o.
Doosan
eMyslivost.cz
Smart work system s.r.o.
Buty
More iQuest referencesshow more

Working on a company information system?

Tell us which agenda is holding you back most. In the first consultation we go through the process, the existing systems and the data, and tell you whether an extension of what you have will do or whether the system is worth rebuilding — including indicative scope and price.