The increasing demand for semantic interoperability and regulatory compliance in the financial sector has led to the growing interest in constructing financial knowledge graphs (KGs). However, ontology alignment and personal data detection remain two significant challenges for transforming financial datasets into machine-readable, privacy-aware knowledge graphs. In this paper, we present an automated pipeline that integrates Large Language Models (LLMs) for aligning tabular financial data to the Financial Industry Business Ontology (FIBO) and detecting personal identifiable information (PII) using the Data Privacy Vocabulary (DPV). The system processes raw CSV data, performs schema-level and instance-level semantic enrichment, and generates privacy-aware RDF graphs suitable for advanced analytics and regulatory reporting. Our approach leverages retrieval-augmented generation (RAG) and ontology-guided prompting to enhance alignment accuracy and privacy annotations. We demonstrate the applicability of the proposed method on representative financial datasets, showing its ability to produce high-quality, semantically enriched knowledge graphs while automatically identifying sensitive data. The proposed solution improves the efficiency, scalability, and compliance-readiness of knowledge graph construction processes in the financial domain.