4. Implement Information Extraction Solutions (10-15%)

This domain is lighter by weight but central by gravity: it's where grounding (Phase 1) and agent knowledge sources (Phase 2) become a concrete pipeline. Information extraction means turning your documents and data into something a model can retrieve from and reason over — chunking, embedding, indexing, retrieving, and grounding with citations. Because RAG underpins so many agent scenarios, expect these concepts to surface in questions nominally about other domains, not just here.