- Develop Semantic Mapping Models using LLMs to match field descriptions even when labels differ (e.g., mapping \”Last Traded Price\” to PX_LAST).
- Implement Fuzzy Matching Algorithms (Levenshtein, Jaro-Winkler, Cosine Similarity) to handle near-matches in text strings.
- Build Vector Embeddings of the Vendor Data Dictionary to enable high-speed semantic search.
- Design a \”Confidence Score\” framework to flag low-certainty matches for human review.
AI/ML Skills:
- LLMs & NLP: Experience with OpenAI API, LangChain, or HuggingFace Transformers for text-to-field understanding.
- Vector Databases: Proficiency in Pinecone, Milvus, or Weaviate for managing high-dimensional data embeddings.
- Automation: Building \”Human-in-the-Loop\” (HITL) workflows to retrain models based on SME feedback.
#J-18808-Ljbffr
Contact Detail:
Ubique Systems Recruiting Team