Machine Learning Resident - Client: Ultimarii (12 month term)
Core
Researching, evaluating, and benchmarking graph retrieval strategies to enhance a hybrid RAG pipeline with a GraphRAG layer for regulatory and compliance challenges in the legal landscape.
Role type
Machine Learning Resident (Research & Applied NLP)
Builds
Graph-augmented retrieval pipelines and domain-specific knowledge graphs
Domain
Legal technology / Regulatory compliance / Knowledge Graphs
Deliverable
production ML models
Required skills
LLMs, Natural Language Processing (NLP), Knowledge Graphs, Graph Retrieval, Entity Extraction, Relation Extraction, Python, PyTorch, Neo4j, LangChain, LlamaIndex
Preferred skills
Prompt engineering, Fine-tuning, Retrieval pipeline optimization, Academic publication record
Technologies
Neo4j, NetworkX, PyTorch Geometric, spaCy, HuggingFace, Transformers, RDF, OWL, SPARQL
Responsibilities
Integrate a graph knowledge layer into an existing hybrid RAG pipeline; Build and populate domain-specific knowledge graphs from unstructured text; Experiment with and evaluate graph retrieval strategies; Build, train, and evaluate features for LLM applications; Collaborate with stakeholders to develop MVP solutions; Prepare presentations and reports on findings.
Seniority
Junior to Mid-level (Residency)