TLDR: Swiss employers hiring for artificial intelligence mostly advertise for data scientists, data engineers and data analysts who make data usable, while prompt writing survives as a skill inside broader jobs.
Swiss artificial intelligence job ads describe data work far more often than model work
The clearest picture of what Swiss employers actually ask for comes from the Lucerne University of Applied Sciences and Arts (HSLU). Its Swiss AI Jobs Report 2025, published on 17 September 2025, analysed 4,703 artificial intelligence (AI) job advertisements posted in Switzerland between July 2024 and June 2025, with demand running above 400 postings a month since early 2025. The postings were supplied by the Swiss job-data firm x28 and classified with machine learning (ML), so the dataset reflects real advertisements.
Six core roles account for 2,774 of the 4,703 postings, about 59%, and three of them are data roles. The full HSLU report counts 767 data scientist postings, 662 data engineer postings and 597 data analyst postings, together roughly 43% of the whole sample. Machine learning engineers, the people who put models into production, appear in 181 postings, under 4% of the total. The AI internship, at 398 postings, is a larger category than the ML engineer.
HSLU names software development, database engineering, cloud computing, statistical analysis, data visualisation and natural language processing as the recurring requirements. For data engineers, the top hard skills are data modelling, data analysis, data integration and pipeline development, and the most requested tools are Azure, Databricks, Spark, Amazon Web Services (AWS) and Docker. In the report’s tool counts, Structured Query Language (SQL) appears in 427 data engineer postings, slightly more often than Python at 408.
| Core artificial intelligence role | Postings | Mid or senior level |
|---|---|---|
| Data scientist | 767 | 60% |
| Data engineer | 662 | 79% |
| Data analyst | 597 | 60% |
| Intern in artificial intelligence | 398 | 2% |
| Machine learning engineer | 181 | 69% |
| Researcher in artificial intelligence | 169 | 59% |
The prompt engineer survives as a skill inside broader jobs
In 2023 the prompt engineer was the emblem of the AI labour market, and Korn Ferry recalls that people with the skill commanded salaries of up to 200,000 US dollars (USD). Two years later the title had largely dissolved. Korn Ferry, summarising a Microsoft survey of 31,000 workers in 31 countries, reported in April 2025 that prompt engineering ranked second from the bottom among the new positions employers planned to add over the following 12 to 18 months, while AI coding, AI data interpretation and AI security led the list. Microsoft’s own 2025 Work Trend Index found that 78% of leaders were considering hiring for AI-specific roles, naming AI trainers, data specialists and security specialists among them.
Allison Shrivastava, an economist at Indeed, told Fortune in May 2025 that prompt engineering “is still definitely a good thing to have” as a skill inside a broader title, adding that searches for prompt engineering roles on Indeed peaked in April 2023. The six core roles that HSLU extracted from 4,703 postings are data scientist, data engineer, data analyst, AI intern, ML engineer and AI researcher; prompt engineer is absent from that list.
A requisition titled “prompt engineer” attracts generalists who are fluent with chat interfaces and leaves the hard problem, clean and governed data flowing into a model, with an unnamed owner. A requisition titled “data engineer, AI platforms” attracts the profile that Swiss employers already compete for, and it describes work that remains valuable whichever model vendor the company chooses next year.
PwC Switzerland’s AI Jobs Barometer 2026, released on 15 June 2026, counts every advertisement that mentions AI skills and finds that Swiss postings requiring AI skills rose by around 9,000 in 2025 to about 25,000, 1.8% of all Swiss postings, with most of the growth in roles where AI supports another profession. The HSLU sample, by contrast, isolates dedicated AI positions, and that is where data roles lead.
Regulated Swiss industries need data engineers who can document what they build
The HSLU report places 298 postings in pharmaceuticals and chemicals, behind information technology (1,205), financial services (552) and research and education (435), and ahead of insurance (252). By region, the Zurich region held 1,773 postings, over 37% of the national total, followed by Geneva (536), Bern (512), Basel-Stadt (448) and Vaud (281). For Edward Galle’s clients along the Lake Geneva and Basel life-sciences corridors, those figures describe a thin, contested pool: the HSLU press release reports that single postings often draw hundreds of applications, that vacancies are frequently closed again within a week, and that selection usually runs over two to four rounds including technical case studies.
In financial services, the Swiss Financial Market Supervisory Authority (FINMA) issued Guidance 08/2024 on 18 December 2024, asking supervised institutions for “appropriate identification, assessment, management and monitoring of the risks resulting from the adoption of AI”, and naming model risks such as weak robustness, correctness or explainability, and bias. In pharmaceutical manufacturing, the European Commission and the Pharmaceutical Inspection Co-operation Scheme (PIC/S), whose participating authorities include Swissmedic, opened a consultation on 7 July 2025 on revised good manufacturing practice (GMP) rules that introduce a new Annex 22 on artificial intelligence. As summarised by PharmTech, the draft keeps generative AI and large language models out of critical GMP applications and requires procedures for human review.
Both instruments point to the same hire. A bank or a manufacturer under inspection needs engineers who can trace a training dataset, version a model, test it against acceptance criteria and write the validation record an inspector will read. That combination of data engineering and computerised-system validation literacy is scarcer than either skill alone, and it belongs in the job description from the first draft. Drug safety shows the same shift: Edward Galle’s article A 2025 European Union Rule Puts Signal Detection Inside Drug Companies in Europe shows how that rule created demand for pharmacovigilance data analysts and safety systems specialists with exactly this profile. Edward Galle’s analysis of AI in regulated talent acquisition covers the compliance side of the same question from the recruiter’s chair.
Language is the second Swiss filter. English dominates most AI advertisements, yet the HSLU language counts show German in 103 ML engineer postings against English in 97, and French in 113 data engineer postings. In Romandie and Basel, a data engineer who can present a validation summary to a French- or German-speaking quality unit carries an advantage that is scarce in an international candidate pool.
Data engineers in regulated industries are priced against the technology market
A data engineer can work for a bank, an insurer, a software firm or a pharmaceutical group with the same toolset, so pay follows the information and communication technology (ICT) market. The swissICT salary study 2025 reported a median base annual salary of 130,060 Swiss francs (CHF) at senior level, the most frequently reported competence level. Edward Galle’s article Digital Hires in Swiss Pharma Are Priced Against the Technology Market, Above the Pharma Median shows why an offer anchored on an internal pharmaceutical grade tends to land below that market.
In the HSLU sample, 56% of data engineer postings asked for mid-level and 23% for senior or lead experience, with 9% at entry level, so most employers compete for the same experienced engineers. Edward Galle’s view is that regulated-industry employers hold an offer of their own against the technology market: a visible link between the engineer’s pipelines and patient safety, product quality or financial stability, plus the validation experience that makes a profile more valuable over time. That argument works best when the base offer is already credible against the technology benchmark.
Hiring managers attract stronger shortlists by writing the requisition around the data problem
Hiring managers in pharma, medical technology and finance get better shortlists when the job description names the data problem first. A strong Swiss brief states the source systems (an enterprise resource planning platform, a laboratory information system, a customer relationship management database), the target platform (Azure, Databricks or AWS, the tools the HSLU data shows candidates already list), the regulatory frame (GMP, FINMA Guidance 08/2024 or the medical-device rules) and the evidence expected at interview. Employers who want to grow talent internally have a real pipeline in the 398 AI internships, where a bachelor’s degree is required in 46% of cases and a master’s degree in 29%.
When an employer plans a graduate route into data engineering, Edward Galle starts from the 9% entry-level share: fewer than one data engineer advertisement in ten is written for a beginner, so a graduate who targets the open market competes for a small set of openings. Read from the employer side, the same figure means an employer who builds its own entry route competes with fewer rivals for the same graduates. The employers who gain most from that gap design the route before they advertise it. They name one production pipeline the graduate will own within the first year, pair the graduate with a senior engineer who signs off each change, and write the validation and documentation steps into the training plan, so the graduate learns the regulated part of the job alongside the technical part. The 398 AI internships in the HSLU sample show where that pipeline can start, and an internship that converts into a junior data engineering post within a defined period gives both sides a clear decision point. Graduates weighing such offers can ask three practical questions at interview: which pipeline they would own, who reviews their work, and when the employer decides on a permanent role.
A vacancy that closes within a week elsewhere is lost during a four-week internal approval cycle, so a hiring team benefits from agreeing the technical case study, the panel and the salary band before the advertisement goes live. Edward Galle’s guide to AI on both sides of the hiring table sets out how screening tools interact with that timeline.
Edward Galle recommends a three-step interview loop for data engineers in regulated Swiss employers, matched to the two to four rounds that HSLU describes. The first conversation covers the data problem and the candidate’s past pipelines, led by the hiring manager. The second is a short technical case study built on a realistic but anonymised extract, such as joining laboratory and manufacturing records or reconciling customer data across two systems, followed by a discussion of how the candidate would test and document the result. The third brings in a quality, compliance or risk partner who asks how the candidate would explain a pipeline change to an inspector. Three rounds, scheduled within ten working days, keep the process inside the window in which strong candidates stay available.
Data and engineering candidates should treat SQL, data modelling and one cloud platform as the entry ticket, and add a short portfolio that shows a pipeline built, tested and documented end to end. In regulated sectors, a written validation note or a model card attached to that portfolio speaks directly to the inspector’s question. Candidates targeting German-speaking or Romand employers should state their working languages in the first lines of the curriculum vitae (CV), because the HSLU counts show the requirement sits inside the advertisement.
Data analysts form the third-largest group in the HSLU sample and offer the most natural route into data engineering. The analysts who make the move fastest already own a reporting layer end to end: they write the SQL behind the dashboard, schedule the refresh and trace the upstream data when a figure looks wrong. Power BI, the tool HSLU lists most often for data analysts, is a common starting point; adding one orchestration tool and one cloud data platform turns that experience into an engineering profile. Employers facing a thin senior pool can treat such analysts as a deliberate internal route, paired with a senior engineer for the first validation cycle, so that documentation habits form from the start.
Marketers, commercial analysts and other domain specialists sit in the wider group that PwC counts, where most of the 2025 growth occurred. For them, the right move is to become the AI-fluent version of their current profession; the European Union’s AI Act, which binds many Swiss exporters, requires providers and deployers to “take measures to support the development of AI literacy of their staff” under Article 4, as replaced by Regulation (EU) 2026/1744 in force since 27 July 2026. Edward Galle’s reskilling roadmap from marketer to brand-tech specialist describes one such path in commercial life sciences.
As of August 2026, the Swiss AI labour market rewards two kinds of people: engineers who make data trustworthy, and domain experts who use AI well inside their own field. Candidates in either group can submit a CV to Edward Galle ou browse current openings, and employers looking for data, brand-tech or revenue-operations profiles in regulated industries can brief the search directly.
This article has been reviewed by Haider Alleg, Managing Partner at Kainjoo SA.
References
- Lucerne University of Applied Sciences and Arts (HSLU). Swiss AI Jobs Report 2025: Neue Studie zeigt Hotspots für KI-Stellen in der Schweiz. 17 September 2025. https://www.hslu.ch/de-ch/hochschule-luzern/ueber-uns/medien/medienmitteilungen/2025/09/17/news-ai-report/
- Lucerne University of Applied Sciences and Arts (HSLU). Swiss AI Jobs Report 2025 (full report). September 2025. https://www.hslu.ch/-/media/campus/common/files/dokumente/h/1-medienmitteilungen-und-news/2025/i/ai-jobs-report-2025-hslu.pdf?sc_lang=de-ch
- Korn Ferry. The Hottest Job of the Year Goes Cold. 29 April 2025. https://www.kornferry.com/insights/this-week-in-leadership/the-hottest-job-of-the-year-goes-cold
- Microsoft WorkLab. 2025 Work Trend Index: The Year the Frontier Firm Is Born. April 2025. https://www.microsoft.com/en-us/worklab/work-trend-index/2025-the-year-the-frontier-firm-is-born
- Fortune. Prompt engineering was a $200k six-figure role. Now it’s obsolete thanks to AI. 7 May 2025. https://www.fortune.com/2025/05/07/prompt-engineering-200k-six-figure-role-now-obsolete-thanks-to-ai
- PwC Switzerland. AI Jobs Barometer 2026 (press release). 15 June 2026. https://www.pwc.ch/de/presse/ai-jobs-barometer-2026.html
- swissICT. Saläre der ICT 2025: Moderate Lohnentwicklung und vertiefte Einblicke in Anstellungsbedingungen. 2 September 2025. https://www.swissict.ch/salaere-der-ict-2025/
- FINMA. FINMA Guidance 08/2024: Governance and risk management when using artificial intelligence. 18 December 2024. https://www.finma.ch/en/news/2024/12/20241218-mm-finma-am-08-24/
- PIC/S. Joint stakeholders’ consultation on the revision of Chapter 4, Annex 11 and the new Annex 22 (Artificial Intelligence). 7 July 2025. https://picscheme.org/en/news/joint-stakeholders-consultation-on-the-revision-of-chapter-4
- PharmTech. Stakeholders’ Consultation on GMP Annex 11, Chapter 4, and New Annex 22. 22 August 2025. https://www.pharmtech.com/view/stakeholders-consultation-on-gmp-annex-11-chapter-4-and-new-annex-22
- European Union. Regulation (EU) 2024/1689 (Artificial Intelligence Act). Official Journal, 12 July 2024. https://eur-lex.europa.eu/eli/reg/2024/1689/oj
- European Union. Regulation (EU) 2026/1744 amending Regulation (EU) 2024/1689 (Digital Omnibus on AI), Article 1(5). Official Journal, 24 July 2026. https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=OJ:L_202601744
