Freelancer profile translated to English.
Back to original languageAbout Deschances
Senior Data & AI Architect with over ten years of experience designing and deploying cloud-native data platforms and AI solutions for large organizations in the banking, insurance, and automotive sectors.
Specializing in the architecture of secure and compliant AWS and Cloudera systems — data lakes, data mesh, real-time streaming, and GenAI/RAG solutions — capable of scaling from terabytes to petabytes while adhering to cross-border regulatory requirements (MiFID II, GDPR, HIPAA, IFRS, Solvency II).
Key achievements: migrating over 100 pipelines to AWS for Crédit Agricole CIB, integrating 6+ business domains into federated data products at Bosch, optimizing Spark workloads at BMW, designing agentic RAG architectures with Amazon Bedrock, and deploying multi-region CI/CD pipelines via Infrastructure as Code across Europe and North America.
Project types: data lake construction and modernization, data mesh architecture, real-time pipelines, cloud migration (Hadoop to AWS), GenAI/RAG platforms, Data Vault 2.0, ELT with dbt/Redshift, and cloud cost optimization.
French
Native or bilingual
Italian
Native or bilingual
English
Fluent
Can work on-site
Paris (up to 20km), Lyon (up to 10km)
Experience
- ODD ParrotSenior Data ArchitectDIGITAL AND ITOctober 2025 - Today (8 months)Zurich, SwitzerlandIndependent Data & AI Architect, designing and deploying production data platforms, domain data products, and GenAI solutions for clients in financial services.UBS Global Wealth ManagementParticipated in a global data mesh transformation program (26 countries).- Drafted domain-aligned data product standards, data contracts, and federated patterns enabling international reuse.- Specified cross-border data access policies to meet European, APAC, and Swiss regulatory requirements.**Stack**: AWS, Python, TerraformLucidoc.AIDesigned an agentic RAG architecture for an AI platform on sensitive content.- Designed the retrieval orchestration, evaluators, and security guardrails with Amazon Bedrock (Claude).- Developed ETL pipelines for ingestion, chunking, embeddings, and vector search, with IaC-driven CI/CD and observability.**Stack**: Amazon Bedrock, Lambda, S3, Python, React, Terraform
- Amazon Web ServicesSenior Data architectDIGITAL AND ITFebruary 2020 - September 2025 (5 years and 7 months)Zürich, SwitzerlandData & AI Architect within AWS Professional Services, I supported major European clients in their data transformation programs by designing scalable cloud-native architectures and leading Hadoop to AWS migrations.**Crédit Agricole CIB**: led the data lake migration (100+ pipelines, petabytes), resilient EMR/Kafka/EKS/HBase architecture, banking compliance.**Stack**: EMR, EKS, Redshift, Kafka, Spark, Python, Terraform, TableauStandard Bank CIB: multi-account data mesh solution for investment banking (risk, trading, MiFID II compliance).**Stack**: AWS Glue, Spark, S3, Apache Iceberg, Terraform**Roche**: Data Vault 2.0 architecture for the enterprise data warehouse, ELT pipelines with dbt/Redshift, improved data lineage for regulated scientific workflows.**Stack**: Python, Redshift, dbt, SQL**BMW**: optimized critical Spark workloads on the enterprise analytics platform.**Stack**: AWS Glue, Spark, S3, Quicksight, Terraform**Bosch**: data mesh architecture for 6 business domains, onboarded in 2 months. Reusable data models and federated governance.**Stack**: Python, AWS Glue, Lake Formation, Terraform**Amazon.com**: migrated logistics simulation workloads to AWS, led a team of 4 engineers, multi-region CI/CD pipeline.**Stack**: Java, AWS Batch, EC2, S3, CDK
- CREDIT SUISSEBig Data Platform EngineerBANKING AND INSURANCEMarch 2018 - January 2020 (1 year and 10 months)Zürich, SwitzerlandBig Data Engineer responsible for the scaling, maintenance, and high availability of a Cloudera Big Data platform, intended to support banking compliance and regulatory workloads.**Key Achievements**:- Extended the Cloudera cluster from 50 to 90 nodes to support a growing load of regulatory and compliance workloads.- Automated cluster configuration management using Salt (Cloudera Stack, Anaconda, specific toolchains).- Developed Python and Shell scripts to automate node onboarding via Cloudera Manager.- Implemented a Splunk-based monitoring system: daily Python script for detecting new file deliveries, dashboards for platform teams and project managers.- Ensured platform stability with an average incident rate of 1 per quarter and a recovery time of 3 hours.**Stack**: Cloudera, Hive, Impala, Sentry, Kerberos, Python, Shell, Splunk, Salt, Tableau
Recommendations
Be the first to recommend Deschances
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
- Master in Complex Systems ArchitectureEcole Polytechnique, Paris2015Master in Complex Systems Architecture
- Computer Science EngineeringPolitecnico di Torino Turin2013Computer Science Engineering
Certifications
- AWS Certified Solutions Architect – ProfessionalAmazon Web Services Training and Certification2025
- AWS Certified Machine Learning – SpecialtyAmazon Web Services Training and Certification2024