One house for the whole information system
From process analysis and architecture through development, integration and AI automation to cloud infrastructure and support — we cover the full software lifecycle. The same team that builds and runs our own products builds a system tailored to you.
- Information systems
- Enterprise and ERP
- AI agents and automation
- Mobile and web apps
- Integration and cloud
- Data and BI
A software house with product experience
We don't write code to satisfy a spec. We build systems that run in production, under real traffic and a real audit.
Konis Software builds its own products — from an eCommerce platform and AI agents to fiscalization, ERP, WMS and document management. Every recommendation we make comes from systems we operate ourselves, not from slides.
We offer design and development as a service to the same engineering standard. We start from how data actually travels through the company, then choose technology, architecture and rollout order. Screens and modules come last.
We work by integration, not demolition. Your ERP, accounting and warehouse stay where they are; we connect them, modernize gradually, and hand over the code, data and infrastructure — with no vendor lock-in.
What we do
Nine areas that together cover everything a serious software house delivers — from the first analysis to a system that runs for years.
Information systems design
Architecture and a domain model before the first line of code — so you don't pay for the wrong decision later.
- A map of business processes and data flows across departments
- Domain model, service boundaries and the contracts between them (API contracts)
- Architecture choice — monolith, modules or microservices, with reasoning
- Non-functional requirements: load, availability, auditability, recovery
- A phased technical rollout plan with a risk assessment
Enterprise solutions and business systems
ERP, MRP and business apps shaped around how your company actually works, not someone else's template.
- ERP and MRP modules: procurement, inventory, production and finance
- Approval workflows, roles and access rights
- Connection to accounting and e-fiscalization
- Data migration from existing systems and spreadsheets
- Reports and dashboards for management
AI solutions
Agents, LLM/RAG, forecasting and document AI — over your data, with measured accuracy instead of promises.
- AI agents for sales, support and internal automation
- RAG over your knowledge base, documents and catalog
- Demand, inventory and churn forecasting over historical data
- Document processing: classification, field extraction and validation
- Model evaluation, guardrails and accuracy monitoring in production
AI process automation
Repetitive back-office steps a person does today — with a person kept on the sensitive points.
- Automatic sorting and answering of inbound messages and email
- Data extraction from invoices, contracts and orders
- Orchestrating steps across existing apps with no manual retyping
- Human control on sensitive steps (human-in-the-loop)
- Measured time saved per process, before and after
Mobile applications
Native iOS and Android when performance decides, cross-platform when the deadline decides.
- Native iOS (Swift) and Android (Kotlin) for demanding cases
- Cross-platform (React Native, Flutter) for faster time to market
- Offline work, synchronization and push notifications
- App Store and Google Play releases, with CI/CD for builds
- Field apps: scanning, signature and location
Web platforms and headless storefronts
Portals and storefronts with a decoupled front and back end — fast, measurable and accessible.
- Headless storefronts connected to NG Commerce or another backend
- Portals for customers, partners and internal use
- SSR and SEO optimization, load speed and accessibility
- A design system and component library for consistency
- Multi-language and multi-currency support
Systems integration and modernization
A layer that connects what you already have and gradually replaces what holds you back.
- API layers and middleware between the ERP, warehouse and sales channels
- Integrations with payment providers, couriers and marketplaces
- Gradual replacement of legacy systems without downtime (strangler approach)
- Real-time and batch data synchronization
- Repeatable, versioned and documented interfaces
Cloud, DevOps and SRE
Infrastructure that survives real traffic, with monitoring and recovery plans in place before they're needed.
- Architecture on AWS, Google Cloud or Azure sized to the load
- Infrastructure as code (Terraform) and containers (Docker, Kubernetes)
- CI/CD pipelines with automated tests and safe releases
- Monitoring, logs, alerts and recovery plans (backup, DR)
- Cost control and scaling to real traffic
Data engineering and BI
So decisions stand on one verified source of truth, not on ten different spreadsheets.
- Data pipelines (ETL/ELT) and a warehouse for analytics
- Unifying data from sales, warehouse and production
- Dashboards and reports with metrics that are actually tracked
- Data quality: checks, versioning and lineage
- Preparing data for AI models and forecasts
How we work
The same process regardless of engagement model: every phase ends in something tangible, not a presentation.
- 01
Discovery and analysis
We understand the business before we propose a solution.
Workshops with the people who do the work, not only managementA map of processes, bottlenecks and integration pointsDefined goals and success criteria - 02
Architecture and plan
We make the expensive-to-change decisions on paper.
Domain model and proposed architectureTechnology choice with reasoning, not by fashionEstimate of scope, timeline and riskA phased rollout plan - 03
Iterative development
Short cycles with a visible result at the end of each.
Sprints with a working demo at the endCode review and automated tests on every commitProgress visible on a shared board - 04
Integration and QA
The system meets real data and real load.
Connection to existing systems and dataAutomated, integration and load testsSecurity review and access control - 05
Deployment and go-live
We release gradually, with a way back at any moment.
Data migration and a trial run on live dataGradual release with fast rollbackUser training and technical documentation - 06
Support and evolution
The system lives and changes — we watch it and improve it.
Monitoring, alerts and SLA response timesRegular releases with fixes and improvementsA shared roadmap for the period ahead
Engagement models
Four ways to work together. We pick one based on how defined the scope is and how much of a team you already have.
- 01
Fixed-scope project
A clearly defined scope, deliverables and timeline, with a fixed budget. You know what you get and when.
When the scope is clear and the goal is a predictable budget and deadline.
- 02
Dedicated team
A complete team dedicated to your product over the long run, shifting priorities sprint to sprint with you.
When the product grows over time and priorities change often.
- 03
Team extension
Our engineers join your existing team and process, working under your lead.
When you have your own team and process but lack capacity or a skill.
- 04
Maintenance and SLA
A retainer for support, monitoring, fixes and small improvements, with agreed response times.
When the system already runs in production and needs reliable support.
Technology
We pick the tool for the problem, not the other way around. These are the technologies we use in production regularly.
Backend
- Node.js
- TypeScript
- NestJS
- .NET
- Spring Boot
- Go
- Python
Frontend
- React
- Next.js
- Angular
- Vue
- TypeScript
- Tailwind CSS
- Vite
Mobile
- Swift
- SwiftUI
- Kotlin
- Jetpack Compose
- React Native
- Flutter
AI and data
- PostgreSQL
- MongoDB
- Redis
- ClickHouse
- pgvector
- OpenAI
- Anthropic Claude
- LangChain
Cloud and infrastructure
- AWS
- Google Cloud
- Azure
- Docker
- Kubernetes
- Terraform
- GitHub Actions
- Grafana
Integration and messaging
- REST
- GraphQL
- gRPC
- Apache Kafka
- RabbitMQ
- MQTT
- Webhooks
- OAuth 2.0
What you get
The measures you can hold us to.
- 4–12
- Weeks to first releaseWe usually ship the first production release within this range, depending on scope.
- 24/7
- Monitoring and supportMonitoring, alerts and an agreed SLA response time for systems in production.
- 0
- Vendor lock-inThe code, data and infrastructure are yours; you can continue the system yourselves or with another team.
- 1
- Accountable teamOne address for all of it — from architecture through development to support.
The figures shown are typical target ranges that depend on scope and existing systems — not a contractual guarantee.
FAQ
Do you build custom software or only offer ready-made products?
Both. Bespoke design and development is a full service — from architecture to production. Our products (NG Commerce, NG Sara, NG eFiscal and others) are proof of that work and often serve as a ready foundation we adapt, instead of building everything from scratch.
How do you set pricing?
Pricing is on request and depends on the engagement model and scope. A fixed-scope project gets a fixed price, a dedicated team and team extension are billed monthly, and maintenance runs as an SLA retainer. Send a short brief to office@konissoftware.com or +381 66 000 977 and we'll come back with an estimate.
Can you work with our existing systems?
Yes — that's usually how we work. We connect to your ERP, accounting, warehouse and sales channels over APIs. We modernize legacy systems gradually, piece by piece, without downtime and without demanding that everything start from scratch.
Which engagement model should we choose?
If the scope is clear and you want a predictable budget — a fixed-scope project. If the product grows over time — a dedicated team. If you have your own team but lack capacity — team extension. If the system already runs and needs reliable support — maintenance and SLA. If you're unsure, we recommend one based on your situation.
Who owns the code and data?
You do. The code, data and infrastructure are yours from day one — we hand over repositories, documentation and access. We don't lock you in: you can continue the system yourselves or with another team.
How do you approach security and data protection?
Through practice, not promises: least-privilege access, encryption in transit and at rest, separate development and production environments, secret management, and audit trails for sensitive actions. A security review is part of every phase, not an afterthought.
Where is the team and how does communication work?
The team is in Belgrade and works in your time zone and your tools. Progress is visible through a shared board and regular demos at the end of each sprint. We communicate in Serbian and English.
Let's talk about your system
Describe the process that's slowing you down or the system you're planning. We come back with a concrete proposal: which area, which engagement model and in what order — no obligation.