Natale Foata

I designed and developed Marty, an AI-powered SaaS platform to help artisans sell better online. Responsible for the technical vision, architecture, and end-to-end development of: Data & AI • Ethical search engine with dynamic rotation • Matching algorithm based on embeddings (pgvector) and semantic scoring • Automatic assignment: orchestration of contact (calls/callbacks) • AI agents to simplify content creation • Data-driven consulting: recommendations to improve visibility and conversion • Clear dashboard to make data actionable by non-techies Data Engineering & Architecture • PostgreSQL schemas with PostGIS & pgvector • ETL/ELT pipelines for standardization and integration of metrics • Centralized monitoring and logs for observability Infrastructure • GCP + Terraform, Docker containerization, GitLab CI/CD • API security (FastAPI + slowAPI), rate limiting • Back-end performance optimization (reduced response times) Product • Back-end: FastAPI (Python) • Front-end: Next.js, React, TypeScript • MVP delivered in 6 months (a classic mistake 😅), followed by rapid iterations to validate the PMF

I managed the technical part of the industrialisation of a strategic Data Intensive application for shipping lines worldwide. Main achievements : • Pipeline optimisation: Design and improvement of dbt models on Snowflake and PostgreSQL, reducing query times and increasing processing efficiency. Results: faster decisions and reduced operational costs. ⚙️📉 • Improved recommendation engine: Industrialisation of a Constrained Optimisation model in a recommendation API. Impact: better recommendations for the business and reduced development times. 🚢💡 • Efficient containerisation: rationalisation of Docker images, reducing build times and improving deployments. Result: optimised use of resources and substantial savings. 🐳⚡ • Improved monitoring and storage : Reinforcement of the PostgreSQL infrastructure for better monitoring of recommendation results. Results: actionable performance metrics and more informed business decisions. 📈🔍

I have contributed to key projects in MLOps and Data Engineering, optimising processes and providing innovative solutions to meet the company's strategic needs. Major achievements : • MLOps stack: Design and implementation of an MLOps stack, significantly reducing costs and boosting operational efficiency. 💰📉 • Crisis management: Rapid ingestion of critical data during a major incident, ensuring the smooth resumption of operations under time constraints. ⏳💼 • Pipeline optimisation: Implementation of continuous deployment pipelines and maintenance of cloud infrastructure, improving the productivity of the Data team. 🛠️ • Data management: Development and maintenance of DAG Airflow for batch processing, and implementation of real-time data acquisition on GCP. 🔄📈 • ML model industrialisation: overhaul of the MLOps environment to maximise model efficiency and productivity. 💡

Versatile role in Data Engineering, DevOps and Machine Learning in an AWS cloud environment. Key achievements: • Real-time processing: Development of streaming scripts with Kafka, Solace and Redis for real-time data processing. 🌊 • Batch processing pipelines: Designed pipelines with Kedro, Snowflake, PostgreSQL and Cassandra, ensuring smooth data flows. 🔄 • High-performance APIs: Creation of REST APIs with FastAPI to improve connectivity between applications. 🚀 • Data Science and Machine Learning tools : Development of advanced solutions to enrich IoT data and facilitate strategic decision-making. 🔬 • Continuous Deployment : Implementing CI/CD strategies with Gitlab, Docker and Kubernetes for fast and reliable deliveries. 🚢 • Robust monitoring : Implemented monitoring solutions with Prometheus and Grafana to ensure system stability and responsiveness. 📊

I've led a number of Data Engineering projects, providing technical solutions to meet the specific needs of clients in the energy and offshore sectors. Major achievements : • Data cleansing and extraction pipeline: Design of a robust pipeline to process industrial equipment maintenance data, coupled with the creation of a real-time dashboard. → Improved operational insights and facilitated proactive decision-making. 📊 • Maritime fleet optimisation: Development of the backend of a route optimisation solution for a fleet of ships, including management of the application's deployment. → Increased efficiency and reduced logistics costs. 🛳️

As part of the Pricing team, I worked on an initiative to update budget allocations in response to the uncertainties associated with nuclear power plant updates. Key achievements : • Automatic detection algorithms : Development and implementation of a machine learning based algorithm to identify cost overrun invoices. → Significant acceleration of the analysis process. 🚀 • Automation of budget analyses : Transformation of a manual process into an automated solution, enabling accurate and efficient budget updates for nuclear power plant projects. 💹

As a Data Scientist intern, I developed an innovative solution using machine learning to extract key elements from privacy policies. This solution was delivered as a real-time web extension, displaying these elements directly in the browser. Key achievements: • Advanced text analytics: Using machine learning algorithms to extract and structure information from privacy policies. 📊 • Intuitive web extension: Design and delivery of an extension enabling instant visualisation of key points in privacy policies. 🖥️

I did civic service with the Marseille Marine Fire Brigade, taking part in rescue operations and immersing myself in a military environment. 🚒