Junior AI/ML Solutions Engineer

BankUnited, Inc.•Miami Lakes, FL

12h

About The Position

BankUnited is seeking an early‑career Junior AI/ML Solutions Engineer to help design, implement, and support end‑to‑end AI solutions that enhance our financial services, improve customer experiences, and optimize internal operations. In this role you'll contribute to data pipelines, ML workflows, and generative AI applications under the guidance of senior engineers. You will develop hands on skills across the AI solution lifecycle from data preparation and infrastructure design to model development and deployment within a regulated banking environment as well as leveraging Microsoft Copilot agents and Microsoft 365 AI capabilities to optimize internal operations and elevate user productivity.

Requirements

Up to 1.5 years of experience through coursework or internships with ML or generative AI required
Foundational Python skills and exposure to ML frameworks required.
Familiarity with cloud ML platforms required.
Basic understanding of Microsoft 365 ecosystem tools (SharePoint, Teams, Outlook) and Copilot capabilities required.
Basic SQL and data modeling knowledge required.
Understanding of NLP and LLM concepts required.
Awareness of data security and privacy considerations required.

Nice To Haves

Experience (academic or professional) implementing RAG or vector search; familiarity with AI orchestration/agent patterns preferred
Experience with explainability and bias mitigation techniques preferred
Experience integrating Microsoft Graph APIs or automation across Teams and SharePoint preferred
Experience in a regulated industry (especially financial services) preferred
Knowledge of model governance frameworks in financial services and quantitative approaches to risk thresholds preferred.
Exposure to designing benchmarking or controlled testing environments for model comparisons preferred
Understanding of vector databases and semantic search technologies preferred.
Familiarity with cloud architecture principles that favor modular, interchangeable components preferred.
Interest in prompt engineering and LLM fine tuning preferred.

Responsibilities

Contributes features to generative AI applications for document analysis, conversational assistants, and predictive use cases; start with proofs of concept and iterate into production with senior support.
Helps evaluate, fine tune, and integrate large language models (LLMs) into workflows; build prompt templates and lightweight evaluation harnesses
Assists in developing and maintaining ML pipelines on AWS (e.g., SageMaker), focusing on reproducibility and automation
Helps implement data pipelines/ETL to prepare datasets for training and inference using S3/Glue/Athena/Redshift.
Supports and maintain existing AI/ML solutions, troubleshooting issues and implementing enhancements as needed
Supports existing AI/ML solutions through maintenance, issue resolution, and documentation of system behavior and updates
Supports model benchmarking and experiment tracking to assess performance, robustness, and bias; document limitations and known failure modes
Partners with senior engineers and compliance with model risk management expectations and ongoing monitoring practices
Participates in deployments with CI/CD and MLOps practices; assist with alerting, logging, and early risk detection dashboards
Leverages Microsoft Copilot Studio and Microsoft 365 Copilot features to build lightweight automation, copilots, prompts, and AI-driven workflows.
Assists in developing copilots/agents that integrate SharePoint, Teams, Outlook, and Microsoft Graph data sources.
Collaborates with stakeholders to identify practical AI opportunities; help convert requirements into technical tasks and clear documentation
Stays current with advancements in AI/ML.
Adheres to and complies with applicable, federal and state laws, regulations and guidance, including those related to anti-money laundering (i.e. Bank Secrecy Act, US PATRIOT Act, etc.).
Adheres to Bank policies and procedures and completes required training.
Identifies and reports suspicious activity.