About Fatih
AI Evaluation & Compliance | Swiss-Bench Author | EU AI Act + FINMA Expertise
Helvetic AI provides automated AI model evaluation and regulatory compliance assessments for Swiss organizations. We operate Switzerland's only AI benchmarking infrastructure (Swiss-Bench, published on arXiv) and deliver fixed-scope
evaluation reports — not hourly consulting.
What we deliver:
- AI Risk Classification: Determine your AI system's EU AI Act risk level with Swiss regulatory overlay (nDSG + FINMA). Fixed price, 3-5 business days.
- AI Model Selection Advisory: Data-driven recommendation on which AI model performs best for your domain, Swiss languages, and compliance requirements. Based on 30+ pre-evaluated models.
- Local AI Setup Advisory: Personalized hardware/software recommendation for sovereign AI deployment. TCO comparison included.
- AI Model Evaluation Report: Full 5-model benchmark comparison — accuracy, Swiss language quality, domain performance, failure modes. Standardized 25-40 page report.
Credentials: PhD (Political Science), Postdoc Harvard & Cambridge, published Swiss-Bench (arXiv), SOC Analyst background, MSc Computer Science (AI specialization).
Infrastructure: 4-node NVIDIA Blackwell cluster (512 GB unified memory). All evaluations run on Swiss infrastructure. No data leaves Switzerland.
German
Native or bilingual
Turkish
Native or bilingual
English
Fluent
French
Basic
Can work on-site
Bern (up to 50km)
Experience
- SBB CFF FFSAI & Automation in Cyber Security & GovernancePUBLIC SECTORApril 2024 - Today (2 years and 2 months)Berne, BE, SwitzerlandSecurity Operations Center (SOC) Automatisierung & AI Prototyping◦ Applied AI R&D (Prototyping): Konzeption und Entwicklung lokaler LLM-Infrastrukturen zur Analyse sensitiver Daten (Data Privacy).◦ Entwicklung eines Advanced RAG Chatbots, um komplexe GRC-Dokumente (Governance, Risk & Compliance) mittels KI durchsuchbar und interpretierbar zu machen.◦ Einsatz von Ollama und vLLM für On-Premise Inferenz ohne Datenabfluss.◦ Nutzung von Docker für Containerisierung und MLFlow für das Experiment-Tracking der Modelle.◦ Data Engineering: Entwicklung von ETL-Strecken zur Integration heterogener Datenquellen (Splunk, Microsoft XDR, ServiceNow) in ein zentrales Security-Dashboard.◦ Governance & Qualität: Optimierung der "Signal-to-Noise"-Ratio in Sicherheitsdaten. Prinzipien der Datenqualitätssicherung, die direkt auf die Prüfung von Finanzsystemen übertragbar sind.◦ Reporting: Konzeption von Management-Dashboards zur Messung der Effektivität von Sicherheitsmassnahmen (Wirkungsmessung).
- DataIQSenior Data ScientistCONSULTING AND AUDITSJanuary 2024 - April 2024 (3 months)Geneva, Switzerland◦ Predictive Modeling: Entwicklung von Zeitreihenanalysen (ARIMA, LSTM) für Finanztransaktionsdaten und Prognosemodelle.◦ Causal Analysis: Anwendung statistischer Verfahren (A/B Testing, Causal Inference) zur Messung der Effizienz von Massnahmen. Forecasting at Scale: Nutzung von Cloud-Technologien (AWS) und Frameworks wie MetaFlow für skalierbare Vorhersagemodelle.
- University of GenevaBehavioral Data Scientist and Senior ResearcherRESEARCHOctober 2021 - December 2023 (2 years and 3 months)Geneva, SwitzerlandNLP-Pipeline: Aufbau einer automatisierten Pipeline zur Analyse von über 16'000 wissenschaftlichen Abstracts (Text Mining) als Basis für Meta-Analysen. Advanced Analytics: Nutzung von Gradient Boosting (XGBoost) und Python zur Identifikation von Schlüsselfaktoren in komplexen sozio-ökonomischen Datensätzen.
Recommendations
Be the first to recommend Fatih
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- MSc Computer ScienceUniversity of Colorado Boulder2026Computer Science with a specialization in Artificial Intelligence
- PhD Political ScienceHumboldt-Universität zu Berlin2017Quantitative Social Science Research