— 01 INDUSTRIAL
Internal industrial apps
Custom tools for engineering, production, quality processes. Lab management, machine control, shop-floor dashboards. They live inside the client's network, integrated with PLM.
ReactPythonPostgreSQLAzure
Service — 03 / Software Development
Bespoke software,integrated,intelligent.
We build web, desktop and cross-platform mobile applications, picking the right technology for each use case. We integrate heterogeneous enterprise systems and embed AI — conversational chatbots and intelligent features inside the application.
— 01 / Principles
Three core principles.
We build software that speaks the process's language, not the framework's. Technology picked for the case, not the CV.
— 01
Right tech for the case
We evaluate the stack per project: web, desktop, mobile, backend. No arbitrary lock-in, no trendy tech without a reason. The right tool for the right problem.
— 02
Native integration
Software we build lives inside your ecosystem. ERP, PLM, CRM, custom systems — we connect via APIs, middleware and enterprise protocols, no force-fits.
— 03
Embedded AI
RAG chatbots on company knowledge and AI features inside the application: auto-classification, suggestions, semantic search. AI is a feature, not a product on its own.
— Product / SOFTWARE DISTRIBUTION
KitPusher
Push software to your entire PC fleet. In parallel. On-premise. Agentless.
Our enterprise software distribution platform — used internally at FITEC to manage installations and support with a lean team. Parallel push via SMB3 + PsExec, full audit trail, compliance from day one.
- Parallel push via SMB3 + PsExec, on-premise
- AES-256-GCM + PBKDF2 600k iterations, full audit
- From setup to first deploy in 1–3 days
Discover KitPusher
— 02 / Tech stack
The technologies we use, by layer.
We name the specific tools because they matter. Each project picks from this stack what fits — combinations change case by case.
— 01 / LAYER Web
ReactNext.jsAstroTypeScriptAngular
— 02 / LAYER Backend
PythonJava.NETNode.jsPostgreSQLC/C++DockerMongoDB
— 03 / LAYER Desktop
.NET / WPFElectronC++ Qt
— 04 / LAYER AI
Azure OpenAIOpenAI APILangChainMCP (Model Context Protocol)Vector DB (pgvector, Pinecone, Faiss)AnthropicTensorflow
— 03 / How we operate
— 01
Discovery
Understand the real problem. Workshops with end users, flow mapping, identification of technical and organisational constraints. The output is a shared technical brief, not a tender doc.
— 02
Architecture
Solution design: stack choice, data model, API contracts, integration strategy, non-functional requirements (performance, security, scale). Documented and reviewable.
— 03
MVP / Pilot
A first working release in 2–3 weeks on a focused scope. Puts something real into users' hands to iterate against. Drastically cuts the risk of building the wrong thing.
— 04
Build & CI
6–10-week iterations, versioned code, automated tests, CI/CD pipeline to staging and production. Regular client demos, no go-live surprises.
— 05
Operate
Evolutionary maintenance, monitoring, progressive hardening, dependency and security updates.
— 04 / Typical cases
Three recurring project archetypes.
Every client is unique, but these three patterns recur. They're the shapes software most often takes in industrial and technical-scientific contexts.
— 02 AI · MCP
RAG chatbots and MCP servers
Conversational assistants over technical documents (RAG architecture with citations) and custom MCP servers giving models scoped access to ERP, PLM, enterprise databases. AI doesn't just read PDFs — it can query real systems and run constrained actions.
Azure OpenAIMCPLangChainpgvector
— 03 REAL-TIME · BIG DATA
Real-time big-data dashboards
Interactive dashboards integrated with big-data systems: streaming data ingestion, automatic anomaly detection and signalling, live chart visualization. Low-latency backend for high throughput, reactive frontend via WebSocket for instant updates.
Apache SparkReactWebSocket
— 05 / What we build
Six recurring intervention areas.
/ 01
APIs and microservices
Documented REST/GraphQL APIs (OpenAPI), decoupled microservices, versioned contracts. The backbone of any serious project.
/ 02
Containers & deploy
Docker, orchestration (Kubernetes / Azure App Service), Infrastructure-as-Code. Reproducible deploys, identical envs from dev to prod.
/ 03
Async pipelines
Message queues (RabbitMQ, Azure Service Bus) for resilient integrations, retry policy, dead-letter queues, event-driven workflows.
/ 04
CI/CD
Automated build, test, security scan, deploy. GitHub Actions, Azure DevOps. Every commit is a release candidate.
/ 05
Observability
Structured logging, metrics, distributed tracing (OpenTelemetry). Knowing what the software does in production, in real time.
/ 06
Maintainable code
Shared code standards, code review, measured test coverage, continuous refactoring. Software that lasts is software you can change.
— 06 / FAQ
Frequently asked.
— 01 Do you work with technologies we already use in-house?
Yes, it's the default. If you already have a consolidated stack (.NET, Java, Python, specific frameworks) we respect it. We only change it when there's a strong technical reason — not to impose a preference.
— 02 How long is a typical MVP?
Six to ten weeks from kick-off to the first release usable by pilot users. It's not a prototype: it's production code, tested, deployed. From there we iterate on the full scope.
— 03 Is AI really production-ready?
For the right cases, yes. A well-built RAG chatbot with citations and a structured prompt can be a great help in daily use across different fields.
— 03 / Contact
Got an industrial process to digitize?
Tell us the problem. We'll design the right solution together, from proof-of-concept to production.