Thomas Ngo Trung

🧠 AI / GenAI Engineer at SparkWise At SparkWise, I led the design and deployment of advanced agentic AI solutions, leveraging cutting-edge technologies to build intelligent, goal-driven agents for clients. My focus was on creating systems capable of autonomous reasoning, planning, and action. Key responsibilities and achievements: • 🤖 Developed and implemented ReAct agents, Retrieval-Augmented Generation (RAG) agents, and multi-agent systems using LangGraph, Dify and LangFlow, enhancing the autonomy and efficiency of AI workflows. • 📊 Integrated Langfuse for comprehensive observability, enabling detailed tracing and performance monitoring of AI agents to ensure reliability and effectiveness. • ☁️ Deployed scalable AI solutions on Kubernetes, utilizing Azure OpenAI services to provide robust and secure AI capabilities to clients. • 🔌 Implemented the Model Context Protocol (MCP) to standardize interactions between AI models and external tools, enhancing interoperability and reducing integration complexity.

💾 Lead Data Engineer at Continuity: Following a €10M funding round in June, Continuity is experiencing rapid growth. In this phase of rapid growth and structuring, I joined the company. I took charge of upskilling the Data Engineers, supported the founders in defining the data product vision, and prepared RFPs to propose a secure architecture meeting the requirements of insurance clients. My main responsibilities and achievements included: - 👥 Collaboration with the Head of Data to design and scale the data platform, define a clear roadmap, and recruit Data Engineers. - 🧠 Advising leadership on the evolution of the technical architecture, aligning technology with business objectives. - 📚 Establishing best practices for development, DevOps, and DataOps. - 🔄 Migrating the data platform from PySpark (distributed computing) to DuckDB and Polars (out-of-core computing) to reduce complexity, costs, and ETL pipeline execution time. - 🛠️  Developing a data quality framework with data contracts and Pandera, enhancing data reliability. - ☁️ Setting up on-demand development environments based on features via AWS EKS with Fargate, enabling dynamic creation of development environments. - 🚀 Refactoring and optimizing ingestion pipelines with Dagster.

🏛️ Lead Machine Learning Engineer at Allianz Trade: As the Lead Machine Learning Engineer in the Group Credit Assessment team at Allianz Trade, my mission was to implement an ML Ops system for deploying new machine learning models. This initiative was part of migrating their infrastructure to AWS, specifically as part of the Algorithm Management Platform project. My main responsibilities and contributions included: - 👥 Recruitment and management of teams - ✍️ Project management - 🤖 Designing an ML Ops platform with IT teams - 🚀 Developing a micro-service container infrastructure - 🔗 Exposing ML models with FastAPI - 📈 Designing preprocessing pipelines with Pandas - 🔍 Code review and optimization - ⚙️ Implementing dynamic CI/CD with Gitlab CI

💄 Lead Data Engineer for the Beauty Tech project at L'Oréal: As a Lead Data Engineer, I implemented an efficient Row Level Security system and optimized the Data Warehouse under BigQuery for a data science and finance dashboarding project on Looker aimed at L'Oréal's Executive Committee. I was also part of the team responsible for developing the core infrastructure of the data platform. My main responsibilities and achievements included: - 🚀 Optimization of the BigQuery Data Warehouse - 🔐 Implementation of Row Level Security for BigQuery - 📈 Designing data preprocessing pipelines with Dask - ⚙️ Setting up CICD and IaC with Terraform and Github actions - 🤖 Deployment of machine learning models - 🧮 Data modeling for the LookML semantic layer - 🔗 Exposing data to Looker with a FastAPI

🔍 As the Lead Data Engineer at Decidento: I joined Decidento, a B2B data platform, to lead the development of their sourcing solution and exposure of open data from various sources (Insee, Infogreffes, INPI, etc.). My mission covered several key aspects: - ✍️ Defining the architecture - 🔄 Developing open data ingestion pipelines - 💾 Creating PostgreSQL and Elastic Search databases - 🔗 Exposing data via a Python API with FastAPI - 🌐 Implementing Terraform for Infrastructure as Code - 🛠️ Using Github Actions for CI/CD

🌍 At Artefact, an international consulting firm in data and innovation: I was part of the consulting team at the Paris agency, closely collaborating with about 40 data scientists, 12 data engineers, and various business consultants. As a Senior Data Engineer, my main mission was to develop Data platforms and industrialize machine learning projects. My main achievements included: - 👥 Management of the technical data team for projects - 🛠️ Design of several Data Platforms (Marketing, Retail, IoT, etc.) - 🤖 Industrialization of Machine Learning models - 🏭 Development of Data Warehouses and Data Lakes - 🔄 Implementation of CI/CD and Infrastructure As Code

💻 At Azeo, a consulting firm known for its expertise in Microsoft technologies: I joined the Data division as the first Data Scientist in a 20-person team primarily composed of Data Analysts, DBAs, and BI Engineers. My mission: to develop the data science and artificial intelligence offering using Azure cloud. My main achievements included: - 🤖 Development and deployment of Machine Learning models - ☁️ Utilization of Azure's Machine Learning solutions - 📄 Development of an auto-indexing system for documents - 📊 Optimization of distributed computations - 🌊 Development of a datalake - 🛠️ Deployment of Hadoop clusters - 🔍 Code review for data scientists - 📈 Creation of Power BI dashboards

🌟 At Singlespot, a start-up specializing in mobile retargeting and drive-to-store: I joined the team in its early stages, which at that time consisted of the two founders and two mobile developers. My main projects included: - 🔄 Migrating from Heroku to AWS (API and database) - 🛠️ Designing and developing a Data and IoT architecture