Senior AI Full-Stack Engineer

Broadview Federal Credit Union•Albany, NY

About The Position

Broadview is establishing its first dedicated AI function. This role is for someone who can transform AI capabilities into practical products for users, not just pilots or proofs of concept. The goal is to ship production applications into the hands of employees and members securely within weeks, not quarters. There is no existing AI codebase to maintain or inherited backlog; the first project will be entirely new. The role involves working on a small, senior, high-autonomy team alongside the VP of Artificial Intelligence and an AI Automation & Solutions Architect.

Requirements

Bachelor's degree in Computer Science, Information Systems/Technology, Data Science, Math, Electrical Engineering, or Machine Learning (Artificial Intelligence).
7+ years building and shipping production software.
At least 2 years building AI or LLM-based applications that real users depend on.
Genuine full-stack depth: modern TypeScript/React or equivalent on the front end, Python or TypeScript services on the back end, relational and vector data stores, and cloud deployment owned in production.
Hands-on fluency with LLM application patterns: prompt design, structured output, tool use and function calling, retrieval and reranking, agentic orchestration, context management, and cost and latency optimization.
Demonstrated LLM evaluation experience: built test sets, measured output quality, and caught regressions before users did.
Working knowledge of AI-specific security: prompt injection, data exfiltration, data and model poisoning, model access control, and secrets management.
Product instinct: ability to understand user workflows and design adoptable solutions to improve their work and member experience.

Nice To Haves

Experience delivering in a regulated environment (e.g., financial services, healthcare, insurance, government).
Working knowledge on model risk management, third-party risk review, and audit or examination cycles.
Integration experience with core banking, loan origination, or payments platforms.
Professional familiarity with enterprise iPaaS (Workato, MuleSoft, Boomi), Anthropic, and AWS Bedrock.
Fine-tuning, distillation, or self-hosted small-model deployment experience.
Contributions to open-source AI tooling, or public technical writing on LLM application engineering.

Responsibilities

Full-vertical delivery: Build and ship secure production AI applications end-to-end, including assessing data access levels with policy and risk teams, model orchestration, API layer, and the user interface.
Agentic workflows: Design and implement multi-step, tool-using systems for retrieval, reasoning, action, and appropriate handoff to humans, determining when deterministic workflows or a combination of approaches is best.
Retrieval over internal knowledge: Build RAG pipelines using policy manuals, procedures, product documentation, and lending guidelines, with measurable and defensible retrieval quality.
Evaluation before launch: Collaborate with QA teams to ensure all systems undergo rigorous scrutiny for regulatory, policy, and framework compliance before promotion, with every AI feature including a test set, scoring method, and regression suite.
Secure systems integration: Connect securely through documented APIs and the iPaaS layer.
Guardrails as a first-class concern: Implement prompt injection defense, PII handling, model drift and poisoning detection, output validation, human-in-the-loop checkpoints, and comprehensive audit logging.
Judgment about production readiness: Make decisions on what software deserves to be hardened and what should be discarded, and when tools or talent are mature enough for re-evaluation.
Engineering standards for the team: Establish and maintain code review, testing, deployment pipelines, documentation, and secure development practices suitable for a federally regulated financial institution.