I start with the business problem, not the technology stack. I build the cloud data platforms companies run on, and I'm now extending that same foundation into the agentic systems running on top of it, across the wider company ecosystem.
Five-plus years turning business problems into reliable, production-grade data systems — across industry, utilities, and fintech.
I own data platforms end to end: architecture, ingestion, orchestration, CI/CD, and cost. My work sits where engineering rigour meets business impact — I partner closely with product, analysts, and stakeholders to ship data products people actually depend on.
Grounded in aerospace-engineering rigour, with shipped machine-learning foundations. The active focus: the data and infrastructure layer beneath LLM and agentic systems — retrieval pipelines, orchestration, and the observability that carries them from prototype to production.
02Impact at a glance
/ FINOPS
30%
Platform cost reduced through warehouse & workload optimization
/ SCALE
20+
Cloud data platforms deployed across business units
/ LEADERSHIP
4
Data engineers led as technical lead
/ TENURE
5+
Years across data-driven industries
03Experience
Agicap
Senior Data Engineer
Paris / Lisbon2023 — Present
Own business-critical data systems end to end — from scoping and architecture to implementation and production support — across batch and near-real-time workflows.
Led platform initiatives spanning CDC, orchestration, ingestion, CI/CD, and infrastructure security, lifting scalability, reliability, and maintainability.
Cut platform costs by 30% through data-warehouse optimization, asset cleanup, workload tuning, and cost observability.
Partner with analysts, product, and business to ship robust data products, marts, alerts, and integrations for operational and decision-making use cases.
Veolia
Lead Data Engineer
Paris2021 — 2023
Spearheaded a Data-Platform-as-a-Service initiative, deploying 20+ cloud data platforms across Veolia's business units.
Designed and maintained platform infrastructure and CI/CD pipelines for scalable, repeatable deployments.
Built transversal capabilities — alerting, connectors, usage dashboards — that raised reliability and adoption.
Managed a team of 4 engineers and advised stakeholders on architecture, workflow design, and data modeling.
AKKA Technologies
Data Architect
Toulouse2020 — 2021
Designed an Industry 4.0 big-data platform for industrial IoT, with batch and streaming in a lambda architecture (load-tested at ~1 MB/s).
Implemented real-time anomaly detection using supervised and unsupervised machine-learning methods.
Delivered a horizontally scalable system serving real operational use cases — alerting, real-time and batch analytics.
04Projects
Independent · Shopify → Pipedrive
Awake — sales automation integration
Designed and deployed automation connecting Awake's Shopify storefront to their Pipedrive CRM — syncing orders and line items straight into the sales pipeline. Killed the manual data entry and kept commercial data in lockstep, so the sales team works from a single source of truth.
Shopify APIPipedriveWebhooksOAuthAutomation
@ Agicap · Fintech
Change-data-capture ingestion
Deployed CDC streams from PostgreSQL to Google Cloud Storage using Airbyte, unlocking near-real-time data flows for business-critical systems.
AirbytePostgreSQLCDCGCSStreaming
@ Agicap · Fintech
Platform FinOps & cost optimization
Led a sustained cost-reduction effort across the data platform — optimizing dbt models, tuning orchestration workflows, and fine-tuning cluster scaling — cutting platform spend by 30%. Separately halved a streaming reporting instance through targeted optimization, cutting its running cost by roughly half.
dbtBigQueryAirflowCluster tuningCost
@ Agicap · Fintech
Churn Me & Downsell Me
Managed Admin APIs that automate company- and client-management processes, and exposed them to HubSpot — so business teams trigger and run these operations directly from their CRM.
Admin APIHubSpotPythonAutomation
@ Agicap · Fintech
Airbyte migration to Kubernetes
Migrated Airbyte from a single-VM deployment to a Kubernetes cluster — moving batch data collection onto scalable, container-orchestrated infrastructure.
AirbyteKubernetesGKEDockerInfrastructure
@ Agicap · Ongoing
AI enablement for analytics
Building an AI context layer over the analytics platform — Looker MCP for natural-language exploration, with structured context across data-engineering repos — so LLM tooling works reliably against real company data.
Looker MCPLLM contextdbtAnalytics
@ Veolia · Enterprise
Data Platform-as-a-Service
Productised a reusable cloud data platform and deployed 20+ instances across Veolia's business units — with shared alerting, connectors, and usage dashboards baked in. One model, many teams, repeatable delivery. Led a team of four to ship it.
TerraformGKECI/CDAirflowLooker
@ AKKA · Industrial IoT
Real-time anomaly detection at scale
Architected an Industry 4.0 big-data platform for industrial IoT — batch and streaming in a lambda design, load-tested at ~1 MB/s — with supervised machine-learning models detecting anomalies live against real operational use cases.
Francisco quickly became a pillar of the team thanks to his adaptability and excellent communication. Beyond his technical skills, he stood out for his energy, initiative, and proactivity.
Arnaud RoletTech Lead Data, VWIS · managed Francisco at Veolia · translated from French
Francisco is a talented engineer who is passionate about his field, with strong technical skills in problem-solving, software design, architecture, and development. His ability to solve complex technical problems has been an asset to our team.
Sadeq Zougari, PhD, PMPEU Project & Data Manager, AKKODIS · managed Francisco
He took the lead on a new Google Cloud Platform product and quickly built the expertise needed to deliver value to the customer. He's an excellent professional and a great person, always enthusiastic and ready for any challenge.
Quentin BaretAlliance Manager, SFEIR · worked with Francisco