BILEL TLILI

Leading the Data System Team (DST) at Christian Dior Couture — owning platform architecture, governance, security & FinOps at enterprise scale. Primary architect for GCP-based data infrastructure, from long-term roadmap to hands-on engineering. 🏗️ Architecture • Designed event-driven, batch & streaming patterns + DST Task Framework (Cloud Run services & jobs for data enrichment, transfer & integration) • Built a custom GCP Execution Node Orchestrator — parallel job processing across multiple GCP projects, routing by criticality, execution time & dependencies. Eliminates saturation + ~5% cost saving. • Pre-configured GCP project templates (Product, Analytics) with embedded IAM & governance controls. ⚙️ DevOps • ~10 Terraform modules: DBT/Dataform deployments, GCP provisioning, project templates & cross-project management. • CI/CD for data pipelines & infrastructure across all environments. 🔐 Governance & Security • Two-mode data access (direct & indirect via authorized datasets/routines) + RLS combining IAM & BigQuery authorized views for per-user filtering at query time. • PAM, WIF (zero-trust), DataGalaxy, automated lineage, RGPD classification & CI/CD compliance controls. 💰 FinOps • 15–25% GCP cost reduction: AI-driven forecasting, BigQuery on-demand pricing optimization & execution node parallelization. • Real-time cost dashboards via Looker & Power BI for proactive budget management. 🤝 GCP Community • GCP chapter: guilds, architectural guidelines & onboarding programs for Data Engineers. • Mentored 7–10 engineers. Architecture roadmaps presented to senior leadership. ✅ Key Achievements → Custom Orchestrator — in prod, smart routing by criticality & dependencies → Terraform Library — ~10 modules, full IaC & pipeline lifecycle → RLS Engine — IAM + BigQuery layer, per-user filtering at query time → FinOps — 15–25% GCP cost reduction → WIF migration (in progress) — zero-trust identity on GCP → Data-Galaxy Sync — automated catalog & documentation sync

Stepped into a full leadership role over a team of 7–10 Data Engineers at Christian Dior Couture — driving technical standards, platform architecture, and security hardening while delivering key data integration and modeling projects across multiple business domains. 🏗️ DBT Architecture — Multi-Team & Multi-Domain • Designed a scalable DBT architecture enabling multiple business domains (Supply Chain, Enterprise Performance Management) to work concurrently on the same project — solving isolation, dependency management & deployment conflicts at scale. • Defined data modeling standards across domains, ensuring consistency and reusability across business units. 🔐 Security & IAM Redesign • Redefined IAM roles across all stakeholders enforcing least privilege principle — reduced attack surface and tightened access boundaries. • Introduced custom IAM roles for external contributors scoped to data ingestion & consumption use cases. 📊 Data Integration — Media & Marketing • Designed and delivered Fivetran integration pipelines for Meta, Google Ads, Snap, LinkedIn & Twitter. • Built the Media Dashboard — full ingestion via Fivetran + Dataform transformation (3-layer: staging → DWH → data mart) for marketing analytics. ⚙️ Platform & DevOps • Improved CI/CD processes, redesigned Git flow for multi-team parallel development. • Implemented automated retro-documentation and cross-service health checks. ✅ Key Achievements → DBT multi-team architecture — concurrent business domains, no conflicts → Least privilege IAM redesign + custom roles for external contributors → Media Dashboard — end-to-end Fivetran to Dataform 3-layer pipeline → Automated retro-documentation — continuous catalog accuracy → Led onboarding & recruitment for a 7–10 person Data Engineering team

Data Engineer on the Dior Data Platform — combining design and hands-on development to build scalable ingestion and transformation pipelines across Retail, Sales & Product domains. Contributed to key architectural decisions including the migration to event-driven architecture and the first FinOps & performance optimization foundations. 🏗️ Architecture & Pipeline Design • Proposed & implemented migration from Cloud Function triggers to Eventarc — event-driven architecture improving decoupling, scalability & maintainability. • Designed 4 generic workflow templates (batch, snapshot, messaging, event-driven) — adopted as platform standards across all domains. • Introduced containerization patterns and task parallelism to optimize performance & reduce cloud costs. 📦 Data Engineering — Multi-Domain • Built end-to-end ingestion & transformation pipelines for Retail & Sales (Xpert, Cegid, Crystal, M3) and Product domains into BigQuery. • Used DBT + GCP services (GCS, Cloud Run, Cloud Workflows) for structured, reliable data processing. • Extended platform scalability via Azure Pipelines. 📊 Monitoring, Observability & Early FinOps • Built sanity check dashboards on Looker covering quota monitoring, alerting & platform health reporting. • Implemented monitoring & alerting framework — proactive detection of processing failures across all pipelines. • Laid first FinOps foundations: cost-aware design, task parallelism & performance-first architecture thinking. ✅ Key Achievements → Eventarc migration — proposed & implemented, modernized platform processing model → 4 generic workflow templates — platform-wide standards → Looker sanity check dashboards — quota, alerting & health monitoring → First FinOps & parallelism foundations — performance & cost optimization → Multi-domain pipelines — Retail, Sales & Product centralized in BigQuery