Saloid is a data and AI consultancy near Munich offering four core disciplines — Applied AI, Data Engineering & Analytics, Cloud Architecture, and Business Intelligence & Strategy. We lead every engagement ourselves, directly.

Four disciplines. One team. Zero hand-offs.

Every engagement is us — Juri and Oleks — from kickoff to handoff. No account managers, no junior bench, no re-learning your context every sprint.

Last updated: April 2026

Saloid's Applied AI & Custom Solutions practice delivers production-ready AI systems — agents, RAG pipelines, and MCP integrations — for enterprises in the DACH region, with EU AI Act compliance from the start.

Applied AI & Custom Solutions

Problem

You've seen the demos. Maybe you've built a few. But getting AI from notebook to production — with reliability, security, and EU compliance — is where most teams stall.

What we do

We build production AI: autonomous agents, RAG pipelines, MCP server integrations, and the MLOps scaffolding to keep them running. Every project includes EU AI Act risk classification so you're compliant before the August 2026 deadline, not after.

Tools

LLMsClaudeGPTopen-sourceLangChainMCPRAGvector storesNode.jsPythonReactMLOps

Outcome

Production AI in under 90 days. 15–20% productivity lift. EU AI Act compliant by design.

Frequently asked

How long does it take to get an AI agent into production? +

We deliver production-ready agents in under 90 days — architecture, dev, testing, MLOps, and EU AI Act classification. Most projects have a working internal release within 6 weeks.

Do you handle EU AI Act compliance? +

Yes. Every project includes risk classification under Article 6. We build the documentation, audit trails, and human oversight before the August 2026 deadline. Juri handles compliance, Oleks builds the technical controls.

What is MCP and why does it matter? +

MCP (Model Context Protocol) lets AI agents connect to your business tools — CRMs, databases, internal apps — without custom code per integration. We build MCP servers so your AI reads and acts on real business data.

Talk to us about your AI project →

Saloid's Data Engineering & Analytics practice builds server-side tracking, consent implementations (OneTrust, Usercentrics), ETL/ELT pipelines, and data quality frameworks that are GDPR-native and audit-ready — without third-party scripts or cookie banners.

Data Engineering & Analytics

Problem

Your tracking is leaking data. Your pipelines break on Mondays. Your analysts don't trust the warehouse.

What we do

We design server-side tracking (GA4/SST without Google's client-side bloat), consent implementations (OneTrust, Usercentrics), tagging architectures, ETL/ELT pipelines, and data quality frameworks that pass audits. Everything privacy-first. No third-party scripts phoning home.

Tools

GTM / Server-side GTMGA4 SSTOneTrustUsercentricsFunnel.ioSparkAWS GlueSnowflakedbtGreat ExpectationsAirflow

Outcome

70% faster time-to-insight. Audit-grade data quality. Zero cookie banners.

See how we rolled out GA360 across 120+ properties →

Frequently asked

What is server-side tracking and why is it better for GDPR? +

Server-side tracking runs analytics on your own server instead of Google's JavaScript in visitors' browsers. We set up GA4 server-side on EU infrastructure — data stays on EU soil, no cookie consent needed for analytics, pages load faster.

Can you fix existing pipelines without rebuilding? +

Usually. We audit, find the failure points — bad joins, missing tests, schema drift — and fix in place. Full rebuilds are a last resort. Most fixes take 4–8 weeks.

How do you ensure data quality? +

Automated checks at every stage — dbt tests, Great Expectations assertions, source freshness checks. When something breaks, you know in minutes, not when a VP asks.

Fix your data pipeline →

Saloid's Cloud Architecture practice architects and migrates data platforms onto EU-sovereign infrastructure — AWS EU, GCP EU, Hetzner — with IaC, cost controls, and zero CLOUD Act exposure.

Cloud Architecture

Problem

Your data sits on US-controlled infrastructure and someone just asked about CLOUD Act exposure. Or your cloud bill tripled and nobody knows why.

What we do

We architect and migrate data infrastructure on EU-sovereign cloud. AWS EU, GCP EU, Hetzner, OVH — whatever fits. Every stack is IaC-managed, cost-tagged from day one, and designed so you're never locked in. We also do sovereignty assessments when teams realise "Frankfurt region" isn't the same as "sovereign."

Tools

AWSGCPHetznerSnowflakeBigQueryTerraformPulumidbtApache IcebergKubernetes

Outcome

Production-ready in under 8 weeks. Up to 30% cloud cost reduction. Full EU data residency.

Frequently asked

Does AWS Frankfurt = EU-sovereign? +

Not necessarily. AWS is US-owned, subject to the CLOUD Act. True sovereignty needs either a non-US provider (Hetzner, OVH) or specific controls like customer-managed encryption keys. We assess your risk and architect accordingly.

How much can you cut cloud costs? +

Usually 20–30%. Right-sizing, killing unused resources, reserved capacity, fixing egress-heavy architectures. We cost-tag from day one so you see exactly what spends where.

What does Infrastructure as Code mean practically? +

Your entire cloud is defined in versioned code (Terraform/Pulumi), not clicked in a console. Rebuild from scratch in minutes. Review changes before they go live. The next person reads the code instead of guessing.

Architect your EU cloud →

Saloid's Business Intelligence & Strategy practice builds KPI frameworks, semantic layers, and reporting in Power BI and Looker Studio that create a single source of truth — privacy-first for DSGVO.

Business Intelligence & Strategy

Problem

You have 47 dashboards and nobody agrees on what "revenue" means. Decisions happen in someone's spreadsheet at 11pm.

What we do

We start with your business model, not your tools. KPI trees, semantic layers, BI systems where numbers are single-source, real-time, and understood by everyone from intern to CFO. Privacy-first — no user-level tracking that violates DSGVO.

Tools

Power BILooker Studiodbt semantic layerKPI frameworks

Outcome

65% shorter decision cycles. One version of the truth.

See how we unified SAP BW reporting for a global supplier →

See how we built marketing analytics for an FMCG brand →

Frequently asked

Power BI or Looker Studio? +

Power BI if you're in Microsoft-land, need complex DAX, or want row-level security. Looker Studio if you're on GCP/BigQuery and need lightweight dashboards. We've shipped both. We recommend based on your stack, not vendor loyalty.

How do you align KPIs across departments? +

KPI tree — a hierarchy from board targets to team inputs. Stakeholders agree before a single dashboard gets built. This is the unglamorous work that makes BI useful: getting finance, sales, and ops to agree on definitions.

Can you work with our existing setup? +

We audit first. Solid foundation → fix and extend. Messy data model → usually faster to rebuild the model and migrate reports.

Align your metrics →

Let's figure out if we're a fit.

20 minutes. No slides. We'll tell you what we'd do, what it costs, and whether you actually need us.

Book a 20-minute call Send a message