Precise profile selection to meet clients' specific needs is a major challenge for consulting firms. This paper introduces TalanSeeker, an innovative system based on the RAG (Retrieval- Augmented Generation) architecture, designed to optimize this process. The RAG architecture combines natural language generation with the retrieval of relevant information, allowing for a finer match between consultants' skills and clients' requirements. We explore the effectiveness of TalanSeeker in the context of Talan, evaluating its ability to improve the relevance of selected profiles and reduce response time. Preliminary results suggest that using the RAG architecture can significantly enhance profile selection processes in consulting, paving the way for new AI-based approaches in recruitment.