Expertise | ProCoders

// how we think How We Think

Architecture before code: we design the system before writing a line
Business logic first: we understand why before we decide how
Test-driven development: automated test coverage on both backend and frontend
Measured, not believed: AI is tested before release and traced in production like any other system
We own the outcome, not just the ticket, embedded in your operations for the long term
Built to last: maintainable, documented, ready for whoever runs it next

// industries & markets Industries & Markets

Telecommunications

Our primary domain: BSS, sales, order management, service fulfilment and provisioning, field service management, billing, and self-care platforms. Managed engineering service running across 8 markets with millions of active subscribers.

Enterprise & SAP

Deep SAP expertise across S4/HANA, CRM, BRIM/RMCA/SOM, and Archive Link. Complex integrations between SAP and the systems that depend on it. Where standard SAP stops, we continue.

AI & Automation

Agentic systems, RAG, document AI, speech and scoring, built into the CRM, billing, contact centre and messaging channels you already run. Model-agnostic and measured per conversation. Production systems, not prototypes.

E-commerce

SAP Commerce Cloud (Hybris) and Shopware delivery for European markets. Custom storefronts, plugins, payment integrations, and Dockware DevOps.

Telecom markets: Croatia, Slovenia, Bosnia and Herzegovina, Montenegro, Serbia, Bulgaria, Greece, Italy Beyond telecom: Germany, Switzerland, Italy, Denmark, UK, USA

// technology stack Technology Stack

procoders@dev:~$

$ cat tech-stack.json
{
  "integrations": [
    "SAP S4/HANA",
    "SAP CRM",
    "SAP BRIM",
    "SAP Archive Link",
    "webMethods ESB",
    "Billing Systems",
    "REST APIs",
    "Swagger/OpenAPI"
  ],
  "backend": [
    "Python/FastAPI",
    "Java",
    "ABAP",
    "PHP",
    ".NET"
  ],
  "frontend": [
    "Angular/NX",
    "TypeScript",
    "RxJS",
    "Angular Material",
    "Tailwind CSS",
    "Vue.js",
    "SCSS"
  ],
  "architecture": [
    "Solution architecture",
    "Enterprise architecture",
    "BPMN",
    "UML",
    "API-first design"
  ],
  "ai": [
    "Multi-agent orchestration",
    "Tool calling",
    "RAG",
    "pgvector",
    "MCP",
    "Model routing & fallback",
    "Speech pipelines (STT / TTS)",
    "OCR & vision",
    "Evaluation & replay testing",
    "Cost guardrails"
  ],
  "ml": [
    "scikit-learn",
    "XGBoost / LightGBM",
    "SHAP",
    "PyTorch"
  ],
  "e-commerce": [
    "Shopware",
    "Custom plugins & themes",
    "Dockware",
    "SAP Commerce Cloud (Hybris)",
    "Composable Storefront",
    "Payment & logistics APIs"
  ],
  "mobile": [
    "React Native",
    "Flutter",
    "iOS/Swift"
  ],
  "databases": [
    "PostgreSQL",
    "Oracle",
    "Redis",
    "MariaDB"
  ],
  "messaging": [
    "RabbitMQ",
    "Redis Pub/Sub",
    "WebSockets",
    "SSE"
  ],
  "infrastructure & devops": [
    "On-prem / bare metal",
    "Docker",
    "Podman",
    "Nginx",
    "Linux",
    "Jenkins",
    "AWS",
    "GCP",
    "Azure",
    "Terraform",
    "HashiCorp Nomad"
  ],
  "testing": [
    "TDD",
    "Jest",
    "Pytest",
    "Jasmine/Karma",
    "Cypress"
  ]
}

integrations

SAP S4/HANASAP CRMSAP BRIMSAP Archive LinkwebMethods ESBBilling SystemsREST APIsSwagger/OpenAPI

backend

Python/FastAPIJavaABAPPHP.NET

frontend

Angular/NXTypeScriptRxJSAngular MaterialTailwind CSSVue.jsSCSS

architecture

Solution architectureEnterprise architectureBPMNUMLAPI-first design

ai

Multi-agent orchestrationTool callingRAGpgvectorMCPModel routing & fallbackSpeech pipelines (STT / TTS)OCR & visionEvaluation & replay testingCost guardrails

ml

scikit-learnXGBoost / LightGBMSHAPPyTorch

e-commerce

ShopwareCustom plugins & themesDockwareSAP Commerce Cloud (Hybris)Composable StorefrontPayment & logistics APIs

mobile

React NativeFlutteriOS/Swift

databases

PostgreSQLOracleRedisMariaDB

messaging

RabbitMQRedis Pub/SubWebSocketsSSE

infrastructure & devops

On-prem / bare metalDockerPodmanNginxLinuxJenkinsAWSGCPAzureTerraformHashiCorp Nomad

testing

TDDJestPytestJasmine/KarmaCypress

// lab Lab

What we build when nobody asked, so we have the answer before someone asks the question.

production-ready, lab

Voice agents

A telephone agent with two interchangeable voice engines and answers grounded in the operator's documents. Alongside it, a self-hosted voice stack (STT, LLM, TTS, every stage swappable) that runs fully inside the operator's perimeter, with no per-minute cost.

benchmark done

Offline document OCR

Five open-source OCR engines, each in its own container, scored on the same set of identity cards. Nothing leaves the machine. Built for KYC and onboarding flows where a document image cannot go to a third-party API.

in daily use

Agentic development

AI runners as teammates: a runner claims a ticket, writes the code on an isolated branch, runs the tests, checks the result in a real browser and returns a report for review. Nothing merges before a human approves. Thousands of tickets on our own projects so far.

./expertise --list Expertise