May 26, 2026 

  May 26, 2026 

  May 26, 2026 

  May 26, 2026 

  May 26, 2026 

Collaborate with Data Science and Quantitative Analytics teams at Wells Fargo on model fine-tuning, retrieval optimization, model guidance, and inference tuning to improve response quality and consistency. Build proof-of-concept Generative AI solutions using LangGraph, Hugging Face Transformers and Tokenizers, PyTorch, LangChain, and BentoML to prototype domain-specific banking AI workflows, evaluate model behaviour, and support model serving readiness. Develop and optimize Retrieval-Augmented Generation workflows by implementing semantic search with vector databases and designing customized reranking approaches to improve context relevance and answer quality. Conduct model experimentation, evaluation, and optimization to improve generation quality, retrieval accuracy, inference consistency, and domain-specific response performance. Evaluate LLM and reranker performance across high-volume inference workflows using model quality checks, retrieval metrics, and output consistency analysis to ensure precise and stable responses.. Design and integrate prompt-level guardrails using Model Armor, Prompt Guard, and LLaMA Guard 3 to support safe, compliant, and reliable model behaviour in enterprise banking use cases. Engineer embedding strategies and optimize semantic similarity retrieval mechanisms to improve contextual relevance across large-scale structured and unstructured banking datasets.

Job Overview

Collaborate with Data Science and Quantitative Analytics teams at Wells Fargo on model fine-tuning, retrieval optimization, model guidance, and inference tuning to improve response quality and consistency. Build proof-of-concept Generative AI solutions using LangGraph, Hugging Face Transformers and Tokenizers, PyTorch, LangChain, and BentoML to prototype domain-specific banking AI workflows, evaluate model behaviour, and support model serving readiness. Develop and optimize Retrieval-Augmented Generation workflows by implementing semantic search with vector databases and designing customized reranking approaches to improve context relevance and answer quality. Conduct model experimentation, evaluation, and optimization to improve generation quality, retrieval accuracy, inference consistency, and domain-specific response performance. Evaluate LLM and reranker performance across high-volume inference workflows using model quality checks, retrieval metrics, and output consistency analysis to ensure precise and stable responses.. Design and integrate prompt-level guardrails using Model Armor, Prompt Guard, and LLaMA Guard 3 to support safe, compliant, and reliable model behaviour in enterprise banking use cases. Engineer embedding strategies and optimize semantic similarity retrieval mechanisms to improve contextual relevance across large-scale structured and unstructured banking datasets. $129,854.00 /year

Education and Experience Requirements

Bachelor’s degree in Computer Science or related with 4 years of experience.