Senior Data Scientist

📋 Description

• Create robust pipelines for the processing of Electronic Health Records (EHR) and medical literature.

• Establish rigorous multi-stage validation frameworks (Sensitivity/Specificity analysis) to guarantee clinical safety and model dependability.

• Customize Large Language Models using SFT, DPO, or PEFT (LoRA/QLoRA) tailored for specialized medical fields and intricate clinical diagnostic reasoning.

• Design hybrid retrieval systems that integrate vector databases with Knowledge Graphs to eliminate hallucinations and ensure factual accuracy.

• Innovate methods to enhance the transparency of model outputs.

⛳️ Requirements

• In-depth knowledge of Transformer architectures and practical experience in fine-tuning LLMs (Llama 3, Mistral, etc.).

• Practical experience with Knowledge Graphs, Triple-stores, or Graph Databases (Neo4j, ArangoDB) and Graph Neural Networks (GNNs).

• Proficient in LangChain / LlamaIndex and vector search engines (Pinecone, Milvus, or Weaviate).

• Hands-on experience with SHAP, LIME, or custom attention-mapping techniques to enhance model interpretability.

• Strong foundation in statistical validation for high-stakes scenarios and managing imbalanced, messy data.

🏝️ Benefits

• Meaningful social impact: your contributions lead to improved patient outcomes, expedited recovery, and a substantial decrease in diagnostic errors.

• Cutting-edge technology stack: engage at the forefront of AI by merging LLMs with structured Knowledge Graphs (Graph RAG).

• Understand the "why": you will not only create a model but also develop a system that clinicians can rely on due to its transparent logic.