Ubique Systems
Machine Learning Engineer
⭐ - Featured Role | Apply direct with Data Freelance Hub
This role is for a Machine Learning Engineer with a contract length of "unknown" and a pay rate of "unknown." Key skills include LLMs, NLP, Python, and vector databases. Experience with OpenAI API and automation workflows is required.
🌎 - Country
United Kingdom
💱 - Currency
£ GBP
-
💰 - Day rate
Unknown
-
🗓️ - Date
March 7, 2026
🕒 - Duration
Unknown
-
🏝️ - Location
Unknown
-
📄 - Contract
Unknown
-
🔒 - Security
Unknown
-
📍 - Location detailed
Glasgow, Scotland, United Kingdom
-
🧠 - Skills detailed
#API (Application Programming Interface) #Pandas #ML (Machine Learning) #Transformers #"ETL (Extract #Transform #Load)" #Automation #BERT #PyTorch #NLP (Natural Language Processing) #Databases #TensorFlow #Langchain #Python #AI (Artificial Intelligence) #Data Dictionary
Role description
Develop Semantic Mapping Models using LLMs (Large Language Models) 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.
Tools: Python (PyTorch/TensorFlow), Scikit-learn, BERT, Pandas.
Develop Semantic Mapping Models using LLMs (Large Language Models) 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.
Tools: Python (PyTorch/TensorFlow), Scikit-learn, BERT, Pandas.
