Job description

If you are ready to join a company that truly cares about its employees, our members, and our community then you have come to the right place! The Opportunity Broadview is standing up our first dedicated AI function. You are the person who turns AI capability into products people actually use. Not pilots. Not a proof of concept. Production applications, in the hands of employees and members, securely shipped in weeks rather than quarters. There is no legacy AI codebase to maintain and no backlog you inherit. The first thing you build will be the first thing of its kind here. You will work on a small, senior, high-autonomy team alongside the VP of Artificial Intelligence and an AI Automation & Solutions Architect. What You Will Own Full-vertical delivery. Build and ship secure production AI applications end to end: assessing appropriate levels data access in partnership with policies and risk team, model orchestration, API layer, and the interface a human touches. You own the whole stack, not a slice of it. There is opportunity for this team to expand as demand necessitates. Agentic workflows. Design and implement multi-step, tool-using systems that retrieve, reason, act, and hand off to a person when they should. Including knowing when a deterministic workflow is the better answer than probabilistic and when to combine approaches via orchestration layers. Retrieval over internal knowledge. Build RAG pipelines across policy manuals, procedures, product documentation, and lending guidelines, with retrieval quality you can measure and defend rather than assert. Evaluation before launch. In concert with our QA teams, all systems must be put through heavy scrutiny to comply with regulatory, policy, and other frameworks ahead of promotion. Every AI feature ships with a test set, a scoring method, and a regression suite. Secure systems integration. Connect securely through documented APIs and our iPaaS layer. Guardrails as a first-class concern. Prompt injection defense, PII handling, model drift and poisoning detection, output validation, human-in-the-loop checkpoints, and audit logging at level that is more than sufficient to produce artifacts for examination. Judgment about what graduates to production. Agentic tooling means you will produce working software fast. The senior part of this role is deciding what deserves to be hardened and what should be thrown away, further knowing when tools or talent have matured to the level where we can revisit it. Engineering standards for the team. Code review, testing, deployment pipelines, documentation, and secure development practices appropriate to a federally regulated financial institution. What Success Looks Like First 30 days. Environment access, stack fluency, and a working prototype of the first prioritized use case. First 90 days. Three AI applications live in production, with the business staff using it daily through an interface that meets their needs. Evaluation and monitoring in place from day one of release to support forecasted ROI and KPIs being delivered. First year. A portfolio of production AI applications, a documented internal pattern library other engineers can build from, and measurable cycle-time reduction across at least three business processes. What We Are Looking For Minimum Qualifications: 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, including 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 you have 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. You have 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. You can sit with a lending processor for an hour, understand their day, and design something they will actually adopt to improve their quality of work and the overall member experience. Preferred Qualifications: 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. How This Role Differs From the AI Automation & Solutions Architect We are hiring both roles, and candidates should know which one they want. The Engineer builds what does not exist: custom AI applications, agentic systems, interfaces, and the code underneath them. The Architect eliminates friction in what already exists: process discovery, document automation, and integration-led solutions on platforms we already own. The Engineer s instinct is “let s build it properly.” The Architect s is “let s configure it and prove the value before anyone writes code.” Both attack the same problems from opposite ends. If you want to write software to obtain reliable needle moving value from AI solutions, this is the role you want. Compensation: $115,000-$150,000 plus a competitive benefits package Bilingual individuals who are fluent in a second language in addition to English are highly encouraged to apply. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability, status as a protected veteran, or any other status protected by applicable law. Broadview FCU is committed to ensuring individuals with disabilities and/or those who have special needs participate in the workforce and are afforded equal opportunity to apply and compete for jobs. If you would like to contact us regarding the accessibility of our Website or need assistance completing the application process, please contact us at talentacquisition broadviewfcu.com

Requirements

Must-haves

If you are ready to join a company that truly cares about its employees, our members, and our community then you have come to the right place!
The Opportunity Broadview is standing up our first dedicated AI function.
Retrieval over internal knowledge. Build RAG pipelines across policy manuals, procedures, product documentation, and lending guidelines, with retrieval quality you can measure and defend rather than assert.
Secure systems integration. Connect securely through documented APIs and our iPaaS layer.
Engineering standards for the team. Code review, testing, deployment pipelines, documentation, and secure development practices appropriate to a federally regulated financial institution.
What Success Looks Like First 30 days. Environment access, stack fluency, and a working prototype of the first prioritized use case.
First year. A portfolio of production AI applications, a documented internal pattern library other engineers can build from, and measurable cycle-time reduction across at least three business processes.
What We Are Looking For Minimum Qualifications: 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, including 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 you have 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. You have 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. You can sit with a lending processor for an hour, understand their day, and design something they will actually adopt to improve their quality of work and the overall 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.
Compensation: $115,000-$150,000 plus a competitive benefits package
Bilingual individuals who are fluent in a second language in addition to English are highly encouraged to apply.

Application prep

Use the role’s language and application details while tailoring your materials.

Keywords to mirror

aiapimachine learningllmtypescriptreactpythonawsaccessibility

Platform: workday
Effort: long