stefan laux
Software engineerZürich, Switzerland

Clear thinking.
Thoughtful
building.

I’m Stefan. I turn messy enterprise data into AI tools people can rely on—from the models underneath to the way it all gets shipped.

Explore my work
Ideas, taking shape.
Currently building at AI & enterprise data Engineering Project deliveryA little about me
01 / The person behind the codeCuriosity, with direction.

A builder’s mind.
A wider perspective.

I build software at the intersection of AI and enterprise data. At , my work spans the full lifecycle: fine-tuning models, designing retrieval pipelines, setting up GPU infrastructure, and automating how it all gets shipped.

Alongside engineering, I’m beginning a part-time bachelor’s degree in Business Informatics at . It’s drawing me toward the planning and delivery side: how work gets scoped, sequenced, and shipped as a team.

Explore my experience
Portrait of Stefan Laux outdoors
Stefan LauxZürich, CH
02 / Selected workFrom shared systems to personal experiments.

Less noise.
More useful things.

A few things I’ve helped build,
and ideas I’ve made real.

SIXApplied AI, in practice

Market guide search

Helped build an AI chatbot for navigating large volumes of unstructured market guide data, replacing a rigid legacy tool and integrating it into the main customer-facing platform.

Context-aware support

Contributed to an AI tool that drafts support email responses using enriched internal context, turning scattered information into a useful starting point.

Retrieval you can evaluate

Fine-tuning embedding and reranker models with synthetic QA data, supported by a regression evaluation pipeline that checks factual accuracy and tone.

Selected projects & experiments01 — 04
SIX Group wordmark
01

six-webcomponents

Enterprise · Design systems

A shared interface language for enterprise applications. A Web Component library built for consistency, performance, and scale at .

TypeScriptStencilUI / UX
Local Chat browser-based AI conversation interface
02

Local Chat

Experiment · Artificial intelligence

A conversation that stays close to home. An LLM chatbot running locally in the browser with WebLLM.

TypeScriptWebLLMshadcnUI / UX
The voiceassistant speech interface
03

voiceassistant

Experiment · Speech & AI

Exploring a more natural way to interact. A voice-controlled assistant combining browser speech APIs with AI.

Web Speech APIWebLLM
The money-planner budget and expense visualization
04

money-planner

Product · Personal finance

Making the flow of money easier to understand. A personal finance tool for budgeting and visualizing expenses.

FirebaseSankeyChart
More on GitHub
03 / What I bringConnected thinking, across the stack.

The details.
The whole picture.

Good software brings the parts together. From models and infrastructure to interfaces and team delivery.

01AI products & retrievalFrom unstructured to useful.

Retrieval pipelines, embedding and reranker fine-tuning with synthetic QA data, and context-enriched AI tools. Evaluation pipelines use AI-based scoring to regression-test factual accuracy and tone.

KotlinSpring BootLangchain4jRAGEmbeddingsRerankersLLM evaluation
02Infrastructure & deliveryThe systems behind the software.

Linux GPU servers, drivers, and vLLM for self-hosted models. GitLab CI/CD with self-hosted runners, OpenShift deployments through Helm and ArgoCD, firewall rules, and identity integration. Contributed to evaluating Azure for GPU workloads across cost, performance, and failover reliability.

LinuxvLLMOpenShiftHelmArgoCDGitLab CI/CDAzureDockerKubernetes
03Full-stack engineeringA complete view of the product.

Day to day, Kotlin, Spring Boot, Langchain4j, and Angular at . A wider foundation in reusable interfaces, backend services, and web applications—from enterprise components to independent projects.

AngularKotlinSpring BootReactNext.jsTypeScriptWebComponentsJavaPythonNode.jsSQLFirebaseCloudflareGit
04Project coordinationFrom a plan to a shared release.

Increasingly involved in scoping and sequencing team workstreams, balancing infrastructure and product priorities. My part-time bachelor’s degree in Business Informatics at is developing this interest in planning, team delivery, and project management.

ScopingPrioritizationWorkstream planningTeam deliveryBusiness Informatics
04 / Experience & educationDownload CV

Two paths.
One direction.

Business Informatics at .
AI and enterprise software at .
Building a broader perspective, in parallel.

01 / EducationOngoing

Bachelor’s degree

Business
Informatics.

Part-time · Wirtschaftsinformatik

Developing the business and organizational perspective behind technical work, with a growing focus on planning, team delivery, and project management.

2026 — PresentExpected graduation 2030
02 / ExperienceOngoing

AI & enterprise data

Software
Engineering.

· Zürich · Hybrid

Building AI-driven tools for enterprise data, from retrieval and model fine-tuning to evaluation, infrastructure, and deployment. Increasingly involved in scoping and sequencing team workstreams across product and infrastructure priorities.

Aug 2025 — PresentFull-time · Zürich
The foundations

Where it
started.

Aug 2021 — Aug 202503

Apprentice Software Engineer

· Zürich · Hybrid

Built a foundation across frontend, backend, mobile, and object-oriented programming. Delivered applications end to end—including a staff management tool and Spring Boot services on OpenShift—from requirements to release. Developed practical CI/CD, containerization, and machine learning skills while helping plan priorities and timelines.

Education04

Vocational Baccalaureate & Federal Diploma

BMS Zürich · ZLI / TBZ

Technical Berufsmaturität with a focus on mathematics, physics, and English, alongside a Federal Diploma in Computer Science, specializing in Application Development.

Have something in mind?

Good things start
with a conversation.

[email protected]