Andres Vervaecke

• Develop and deploy production-ready multi-agent chatbots in the financial domain, leveraging hierarchical agentic architectures with specialized agents for domain-specific reasoning.

• Integrate retrieval-augmented generation (RAG) agents with monitoring and feedback loops to ensure accuracy, regulatory compliance, and continuous improvement; deliver high-concurrency systems that automate support and significantly reduce operational overhead.

• Co-develop a centralized sales platform for Randstad, unifying insights across regions and business units; build end-to-end data pipelines from database design and ETL to API development and real-time synchronization using Change Data Capture (CDC).

• Design and implement cloud-native platforms on Google Cloud (BigQuery, Dataflow, Pub/Sub, Cloud Functions), enabling scalable, secure, and cost-efficient multi-country operations.

• Serve as Technical SPOC for the ML6–GCP partnership, advising on architecture and best practices.

• Deliver annual AI guest lectures at Ghent University and frequently present at Google events, enhancing ML6's technical visibility and thought leadership.

• Mentor junior engineers, providing guidance on cloud-native AI development, best practices, and productionization.

• Built an internal PowerBI tool to optimize audit scope selection for financial teams.

• Developed an anomaly detection model for fraud identification.

• Presented findings and prototype directly to the client, demonstrating business impact.