In the context of open innovation, technology scouting has become a critical activity for identifying strategic partnerships and emerging technological solutions. Conventional keyword-based search mechanisms used in most digital innovation platforms are inherently limited in their ability to capture the semantic complexity of innovation needs and offerings. This paper presents a semantic search engine integrated within a Digital Innovation Platform to support intelligent technology scouting and recommendation tasks. The proposed approach leverages transformer-based language models to encode natural language descriptions of corporate initiatives and innovation profiles into dense semantic embeddings to enable retrieval based on contextual similarity rather than lexical overlap. A case study in the domain of application modernization demonstrates the effectiveness of the semantic matchmaking engine in generating accurate and strategically valuable recommendations.