Skip to content
Oli!Say hello ↗
Oliver Baer / LüneburgSoftware that works on Monday

I buildserioussoftware.

Founder. Architect. Engineer. Still curious.

From medical systems to AI infrastructure, I turn complicated product ideas into dependable tools people can actually use. Clear decisions. Working software. No mysticism required.

Portrait of Oliver Baer
28 years / still shipping
Hello from
Lüneburg
ARCHITECTURE ★ PRODUCT ★ DEVOPS ★ AI SYSTEMS ★ PHP ★ GO ★ PYTHON ★ TYPESCRIPT ★ POSTGRESQL ★ ARCHITECTURE ★ PRODUCT ★ DEVOPS ★ AI SYSTEMS ★ PHP ★ GO ★ PYTHON ★ TYPESCRIPT ★ POSTGRESQL ★ 

Now / Three seats, one person

What’s
on the
desk?

The seats stack on purpose. Each product supplies a different kind of reality check to the others.

Aristaeus / The Cultivator of Solutions

“What if we solved the problem upstream of the problem?”

A voice in the machine.

Aristaeus carries Oli’s register inside nyxCore: anticipatory innovation, cross-domain synthesis, lateral thinking and product vision — hardened by Ipcha self-check, because a visionary who refuses to be stress-tested is only a pitch deck.

anticipatory-innovationcross-domain-synthesissolution-crystallizationlateral-thinkingproduct-vision

Work.

Selected systems with real users, operational consequences and enough scar tissue to be interesting.

New product / by nyxCore

Etch.

Dialogic learning platform

Turn every question into lasting knowledge.

For organisations whose critical knowledge is scattered across documents, processes and individual experts, ETCH transforms real questions into source-grounded answers, personal learning cards and measurable capability — instead of merely storing knowledge or answering it once.

Open ETCH
01

Ask

Start with a real question, not a course catalogue.

02

Answer

Receive a source-grounded answer from your organisation's knowledge.

03

Etch

Turn the useful part into a personal learning card.

04

Review

Repeat it when spaced-repetition science says it matters.

05

Mastery

Make individual and team capability visible.

Knowledge is not retained because it was available. It is retained because it was used, recalled and reinforced.

01

Etch this

02

My Etches

03

Review

04

Mastery

05

Team Mastery

The arc / Twenty-eight years

Built in the real world.

Medical systems taught precision. Media taught scale. Manufacturing taught consequence. Retail taught the counterexample. AI infrastructure ties those lessons together.

  1. 1998 — 2004

    01 / Medical devices

    Philips Medical Systems. Software for equipment radiologists depend on. Six years of learning that software can have weight.

  2. 2006 — 2007

    02 / Consumer portals

    Surf Media GmbH. High-traffic billing tools and third-party API integrations. The scale year.

  3. 2007 — 2014

    03 / Media intelligence

    Kantar Media GmbH. Zend, Solr, OCR, AWS clusters and seven years of turning noise into signal at industrial scale.

  4. 2013 — 2015

    04 / South America

    Salome Media GmbH, co-founded. Social-media monitoring and product-placement analytics for Spanish-speaking markets.

  5. 2014 — 2020

    05 / Manufacturing

    Managing director at Druckgassysteme Wittler: production, prototypes, 3D printing, quotations and customer relations.

  6. 2021 — 2024

    06 / Back into software

    Go microservices, Kubernetes, exhibition applications, Pimcore and a GraphQL gateway replacing legacy XML.

  7. 2025 — now

    07 / AI-native infrastructure

    Consulting, Q-Records and nyxCore. Three seats, one arc. Cross-domain synthesis as a specialisation.

01

Three agents running in production

Father of three. The longest-running deployment he maintains: three small entities that started stateless, acquired goals he did not specify and are approaching autonomy on their own definitions of correctness.

02

Restaurant kitchens

The composure-under-pressure line in the CV is not a metaphor. Professional kitchens taught ten-things-at-once with no dropped plates — the same mental shape as a production incident at two in the morning.

03

Manufacturing floor

Six years running a manufacturing company. Shipping becomes a habit when broken things stop a physical line and every abstraction eventually meets steel, time and a customer.

The stack / In current use

01 / Languages

  • TypeScript
  • Python
  • Rust
  • Go
  • PHP
  • C / C++
  • JavaScript

02 / Infra · DevOps

  • Linux
  • Docker
  • Kubernetes
  • Traefik · nginx · Apache
  • AWS · Google Cloud · Azure
  • Hetzner · IONOS · on-premise
  • Prometheus · Grafana · Loki

03 / Data · AI

  • PostgreSQL · pgvector · Deepgram
  • Redis · MongoDB
  • Next.js 16
  • tRPC v11
  • LLM routing · BYOK · AI Agent Workflows
  • LLM Red Teaming · LLM Personas
  • Anthropic · OpenAI · Google · Ollama
  • Embeddings · HNSW
  • ML · Audio / Videoprocessing

One practitioner. Whole path.

I work between architecture and the shop floor: from an ambiguous product decision through implementation and into stable production. My stronger skill is often deciding what the product does not need.

If your project needs a twenty-seat consultancy, this is the wrong page. If it needs someone who has shipped across six domains and still answers his own email, this is the right one.