Freelancer profile translated to English.

Description

Hands-on Senior Data Architect, 25 years of experience in designing and delivering large-scale data platforms in complex and regulated environments (banking, defense, telecoms). Core expertise across the entire data spectrum: ingestion, streaming, storage, distributed querying (Trino), graph databases, and governance - on-premises, hybrid, and multi-cloud. Over the past two years, this foundation has been extended to production GenAI: multi-role RAG pipelines, VoIP transcription (STT), LLM inference (vLLM), vector search (OS, ES, Vespa), and graph-retrieval augmented search - in air-gapped environments where cloud shortcuts are not an option. Proven ability to deliver where others only do POCs: from Hadoop platforms to next-generation Lakehouse architectures, consistently bridging technical depth with measurable operational results.

Industry field of expertise

Languages

English
Native or bilingual
French
Native or bilingual

Workplace preferences

Can work on-site

Paris (up to 50km)

Confidentiel
GenAI Architect
DEFENSE AND MILITARY
February 2025 - Today (1 year and 4 months)
Paris, France
• Design and deployment of an on-premises VoIP transcription and meeting summary generation solution (lead role)
• Extension, documentation, and industrialization of a self-RAG MVP into a production multi-role RAG solution, deployed in independent instances by classification level up to restricted level (architect)
• Architecture study: sovereign on-device AI companion, integrated into the Office suite — zero cloud dependency
• Technologies: Cisco CMM, Asterisk, RAG, Mistral, Weaviate, vLLM, Pyannote, Faster-Whisper
Architecture GenAI voip RAG VOIP AI
BNP Paribas
GenAI Architect
BANKING AND INSURANCE
February 2024 - April 2025 (1 year and 2 months)
Paris, France
• Design & deployment in PRD of a 100% on-premises Machine Learning + RAG solution to optimize IT incident management (measured MTTF reduction). (pivot role)
• Creation of a unified graph DB schema integrating multiple CMDBs (assets + IT events) for robust root cause analysis (lead role).
• Evaluation and exploration of Vespa for scalability and advanced vector management (lead role).
• Close collaboration with business and technical teams to deliver scalable and reliable solutions (6k+ unique users/day).
• Documentation of architectures and reusable frameworks.
• Technologies: TigerGraph, Kafka, vLLM, Elasticsearch, Vespa, RAG, LLAMA
RAG Metadata Management MDM IT Architecture graphdb
Credit Agricole
Data Architect
BANKING AND INSURANCE
November 2021 - July 2023 (1 year and 8 months)
Paris, France
• Leadership of the architecture and technical integration of 2 CDP clusters, with migration of use cases from HDP.
• Direction of integration and architecture initiatives for the multi-tenant Dataiku CDP platform (data science in retail banking).
Software Architecture SQL System Integration Big Data CDP