AI Engineer CV: Example and Template
This is how AI Engineers in Switzerland can back up production-grade models, MLOps maturity, and governance knowledge with hard numbers instead of buzzwords.
Currently 27 open AI Engineer positions in Switzerland, across 9 cantons, 3 of them from the last 7 days. vontobel advertises the most.
CV example — two-page template
This template is editable straight away — or upload your existing CV and it is carried into this layout automatically.
What qualification is expected of AI Engineers in Switzerland?
Typically, an MSc from ETH Zurich, EPFL, or a university of applied sciences in computer science, data science, or electrical engineering is expected - complemented by demonstrable production experience.
The classic path leads through a BSc in computer science or data science at a university of applied sciences such as FHNW, ZHAW, HSLU, or OST, followed by an MSc in Data Science, Computer Science, or Computational Science at ETH Zurich, EPFL, or a university of applied sciences. Career changers from physics, mathematics, or electrical engineering are also common; what matters then is demonstrating your connection to software engineering. Write the degree with its formal title, institution, and year, for example 'MSc ETH in Data Science, ETH Zurich, 2017'.
Position further education credibly: a CAS or MAS in Machine Learning, Data Engineering, or AI Management at HSLU, ZHAW, or BFH carries significantly more weight in Switzerland than a long list of online courses. Vendor certifications such as AI-102, Databricks ML Professional, or Google Professional Machine Learning Engineer are useful if they match the target company's cloud platform. For roles in banking, insurance, or pharma, add a governance certificate such as ISO/IEC 42001, since audit readiness is part of the role there.
If you studied abroad, have the level equivalence assessed via Swiss ENIC and cite the reference in your CV. Non-EU/EFTA applicants should state their residence permit directly in the header, i.e. permit B, C, L, or G, since recruiters otherwise fear quota issues. If you still lack a Swiss connection, an internship, a master's thesis with an industry partner, or a project on an Innosuisse-funded initiative can help.
How do you present professional experience and projects as an AI Engineer?
For each role, show which models actually ran in production, for how many users, with what latency, quality, and costs.
First describe the context: team size, industry, data volume, cloud platform, and whether you were responsible for prototypes or production systems. The difference between a 'proof of concept' and a 'production endpoint with an SLA' is the most common follow-up question in Swiss interviews. Then formulate three to four bullet points with numbers: hallucination rate reduced from 12 to 3 percent, MAPE from 18 to 11 percent, inference costs from CHF 4.20 to CHF 0.90 per 1'000 requests.
Clearly separate skills: LLM and RAG work, classic machine learning, data engineering, MLOps, and governance. Name concrete tools along with your role in them, such as MLflow for model registry, Argo CD for deployments, Weaviate for vector search, and Terraform for infrastructure. Avoid progress bars for skills; Swiss hiring managers instead evaluate project evidence, code quality, and whether you supported a model after go-live.
Add a short project section with two to three references: open-source contributions, GitHub profile, publications, or a Kaggle result. The connection to business value is important, i.e. time saved, avoided error costs, or additional revenue in CHF. If confidentiality applies, anonymize the industry ('Swiss health insurer, 1'200 employees') rather than omitting the figures.
Length, photo, and Swiss specifics: what do recruiters pay attention to?
Two pages, a professional photo, and clear information on permit status, start date, and salary expectations are considered standard.
Keep the CV to two pages, or a maximum of three if you have over ten years of experience, and summarize older positions in a single line. A professional portrait photo is still common in Switzerland and is rarely viewed negatively for tech roles; it is not mandatory. Details such as year of birth, nationality or permit status, place of residence, and contact information belong in the header, whereas marital status and religion no longer do.
Ensure complete, gap-free month details, as HR departments cross-check them against employment references. Mention your notice period (often three months) and possible start date, which significantly speeds up the process. If a salary range is requested, base it on the market: entry-level roles are often between CHF 95'000 and CHF 115'000, experienced AI Engineers between CHF 130'000 and CHF 160'000, and senior profiles in the Zurich financial sector above that.
Submit a PDF file with a descriptive filename and use a single-column, ATS-readable layout without text embedded in graphics. Write the dossier in the language of the job posting: German in German-speaking Switzerland, French in Romandy, English for international corporations and startups. Also have employment references, diplomas, and two references ready, and note 'References available on request' in your CV.
Where AI Engineer are hired in Switzerland
How the 27 open positions are spread across the cantons.
Figures as a table
| Canton | postings |
|---|---|
| Zürich | 5 |
| Genf | 3 |
| Waadt | 2 |
| Aargau | 2 |
| Basel-Stadt | 2 |
| Bern | 1 |
| Wallis | 1 |
| Tessin | 1 |
Which languages the postings require
Of 21 postings that state a language — in brackets, those requiring professional level.
Figures as a table
| Language | postings |
|---|---|
| English | 15 (14) |
| German | 8 (8) |
| Italian | 2 (2) |
Who hires AI Engineer in Switzerland
Employers with the most open positions. Staffing agencies are excluded.
Figures as a table
| Employer | postings |
|---|---|
| vontobel | 3 |
| Techyon | 2 |
| RUAG | 1 |
| Lengacher & Partner GmbH | 1 |
| Kraken | 1 |
| Callista | 1 |
Full-time or part-time?
How the positions are advertised.
Figures as a table
| Workload | postings |
|---|---|
| Vollzeit / plein temps | 17 |
| Teilzeit / partiel | 1 |
| Temporär | 2 |
The CV in full
To read through and reuse.
Nadine Brunner
Senior AI Engineer, MSc ETH in Data Science
Senior AI Engineer with 7 years of experience building production LLM and machine learning systems for banks, insurers, and industrial companies in Switzerland. I am responsible for the entire lifecycle, from the data pipeline through fine-tuning and retrieval architecture to monitoring, cost control, and audit readiness. I currently oversee 18 production use cases in Azure Switzerland North and lead the AI inventory in accordance with the EU AI Act and ISO/IEC 42001.
What sets me apart
RAG architecture with measurable answer quality: I build retrieval pipelines with hybrid search, reranking, and chunking strategies, and validate their quality with ground-truth sets and RAGAS metrics rather than demo prompts: reduced hallucination rate from 12 to 3 percent.
MLOps from notebook to SLA: Model training, model registry, canary deployment, and drift monitoring run through CI/CD pipelines with MLflow and Argo CD; the path from experiment to production endpoint takes 4 days instead of 6 weeks.
Data management according to Swiss requirements: Experience with data residency in Swiss cloud regions, pseudonymization under the revDSG, and FINMA outsourcing requirements; three bank projects have passed internal risk review without any need for rework.
Cost transparency per token and prediction: I calculate every use case in CHF per 1'000 requests, combine small language models with caching and quantization, and have reduced an annual budget from CHF 430'000 to CHF 90'000 as a result.
Key achievements
Claims assistant for 240'000 documents. Built a RAG assistant for an insurer's claims processing: indexed 240'000 documents, reduced processing time per case from 14 to 6 minutes, achieving 87 percent user acceptance after 6 months among 210 claims handlers.
Demand forecasting for 1.2 million combinations. Developed a forecasting platform for retail: improved MAPE from 18 to 11 percent, nightly batch scoring of 1.2 million article-store combinations in 35 minutes, reducing inventory losses by CHF 1.4 million per year.
Governance evidence without halting projects. Built an AI inventory of 18 use cases classified by EU AI Act risk categories and prepared for ISO/IEC 42001; external audit resulted in 0 major findings, and the approval process for new use cases was shortened from 8 to 3 weeks.
Experience
Senior AI Engineer — Alpstein Intelligence AG, Zurich, since 03/2022
Twelve-person AI team; production LLM and ML services for insurers and banks, hosted in Azure Switzerland North.
- Built a RAG platform with hybrid search and reranking: reduced average response time from 9 to 1.8 seconds, hallucination rate from 12 to 3 percent
- Reduced inference costs per 1'000 requests through caching, routing, and quantization from CHF 4.20 to CHF 0.90, saving CHF 340'000 annually
- Established an MLOps pipeline with MLflow, Kubernetes, and Argo CD: deployment cycle reduced from 6 weeks to 4 days, 18 models in production with drift monitoring
- Onboarded 3 junior engineers and introduced a code review standard, increasing the share of tested pipelines from 45 to 92 percent
Machine Learning Engineer — Limmat Analytics AG, Zug, 08/2019 - 02/2022
Data platform for retail and logistics clients on Databricks; responsible for forecasting and personalization models.
- Developed demand forecasting using gradient boosting and hierarchical reconciliation: improved MAPE from 18 to 11 percent
- Built a feature store for 2 TB of sales data, enabling feature reuse across 6 models, reducing training runtime from 5 hours to 40 minutes
- Validated a recommendation model in an A/B test with 180'000 users: 4.5 percent higher revenue per session with stable latency under 120 ms
- Introduced a data quality monitoring dashboard, reducing pipeline failures per month from 7 to 1
Data Scientist (Computer Vision) — Vorai Systems GmbH, St. Gallen, 09/2017 - 07/2019
Industrial image processing for medtech suppliers; edge deployment directly at production lines.
- Trained defect detection CNN models on 45'000 annotated images: F1 score of 0.94, reducing scrap rate by 22 percent
- Deployed models across 6 production lines via NVIDIA Jetson, achieving inference time of 40 ms per image at 1'800 parts per hour
- Halved the annotation process using active learning: reduced labeling effort from 320 to 150 hours per model version
- Prepared validation documentation for regulated clients, supporting 2 customer audits without findings
Education
Master of Science ETH, Data Science — ETH Zurich · 2017
Bachelor of Science FH, Computer Science, Data Engineering specialization — FHNW Brugg-Windisch · 2015
CAS, Machine Learning and Data Science — Hochschule Luzern (HSLU) · 2021
Microsoft Certified: Azure AI Engineer Associate (AI-102), 2023 · Databricks Certified Machine Learning Professional, 2022 · ISO/IEC 42001 Lead Implementer (AI Management System), 2024 · AWS Certified Machine Learning - Specialty, 2020