As an AI Engineer , you will be the technical engine behind every AI implementation the company runs, setting up the models, building the safety and reliability infrastructure, and establishing the engineering standards that every future AI project will inherit.This is a greenfield role with high ownership. You will be designing and building the foundational AI platform that Hotwire's business units depend on. You'll partner closely with the Director of AI Implementation and AI Champions embedded in each business unit, translating validated workflow proposals into production-grade AI solutions.Duties / Responsibilities Design and build the core AI platform that connects Hotwire's business applications, data sources, and AI models into reliable, production-grade pipelinesOwn the model deployment layer, configure, version, and maintain LLM endpoints across Azure OpenAI and/or AWS Bedrock with environment isolation (dev / staging / prod)Implement a model abstraction layer (e.g., LiteLLM) to ensure portability across model providers and avoid hard vendor lock-inBuild and maintain an internal AI SDK / shared libraries so that future engineers and CoE projects can bootstrap quickly without reinventing plumbingOwn infrastructure-as-code and CI/CD pipelines for AI services Other duties as required or assigned.Actively participate in Steering Committee reviews, translating technical risk and feasibility into language business leaders understandBuild and enforce input/output security controls for every AI-facing endpoint:PII detection and redaction before data reaches external model APIsPrompt injection detection, pattern-based and embedding-based classifiersContent policy filtering and output moderation for customer-facing AI surfacesRole-based access control to AI capabilities across business unitsPartner with IT Security and Compliance to ensure every AI deployment meets Hotwire's data residency, encryption, and access audit requirementsMaintain a centralized secrets management approach for API keys, model credentials, and third-party integration tokensImplement an LLM evaluation framework that every CoE project must pass before production promotionLLM-as-judge pipelines for automated output quality scoringRegression test suits that protect against model drift when providers update underlying modelsSemantic similarity and coherence metrics for RAG-based applicationsGolden dataset management and versioning for reproducible evalsOwn the eval harness integration into CI/CD, no model change ships without passing eval thresholdsTrack and report quality metrics to the Director and Steering Committee as part of the AI implementation lifecycleBuild operational safety infrastructure around AI services:Rate limiting and token-budget enforcement per business unit and use caseCircuit breakers to prevent downstream cascades when model APIs degradeIteration caps and wall-clock timeouts on agentic workflowsAsync queue management and retry logic for high-volume pipelinesConfigure private endpoints and VNet integration for model APIs to keep data off public internet pathsImplement cost allocation and spend controls so that per-department AI usage is visible and accountableSet up comprehensive tracing and monitoring across all AI services using tools such as LangSmith, LangFuse, or equivalentBuild dashboards that surface latency, error rates, token consumption, quality scores, and cost per workflow, visible to both engineering and business stakeholdersEstablish alerting thresholds and on-call runbooks for AI service degradationMaintain audit logs of all model inputs and outputs for compliance reviewServe as the technical reviewer for AI workflow proposals coming from business unit AI Champions before they reach the Steering CommitteeWrite engineering standards, integration patterns, and runbooks that AI Champions and future engineers can followContribute to vendor evaluations, help assess new AI tooling, model releases, and platform optionsOther duties as required or assigned by supervisor.Minimum Qualifications To perform this job successfully, an individual must be able to perform each essential duty satisfactorily. The requirements listed below are representative of the knowledge, skill, and/or ability required.2-4 years building and operating production LLM applications, not prototypes, not demos, production systems with real users and real SLAs4 years of software engineering experience with a strong bias toward system design and production-grade architectureExpert‑level Python, you write clean, tested, maintainable Python, not just scriptsDeep understanding of API design, microservices patterns, async programming, and distributed system fundamentalsHands‑on experience with CI/CD pipelines, containerization (Docker), and cloud‑native deploymentStrong debugging instincts, you can trace a failure from a user‑facing symptom down to a model API edge caseExperience deploying and managing LLMs on enterprise cloud platforms: Azure OpenAI Service or AWS BedrockBenefits 401K Retirement Plan with Company MatchPaid Vacation, Sick Time, and Additional Holidays (including your Birthday!)Hotwire Service Discounts – for employees who live on a property serviced by Hotwire. Discounted service offerings are provided for high‑speed internet, video service, phone, and security serviceHotwire provides equal employment opportunities to all employees and applicants for employment and prohibits discrimination and harassment of any type without regard to race, color, religion, age, sex, national origin, disability status, genetics, protected veteran status, sexual orientation, gender identity or expression, or any other characteristic protected by federal, state or local laws.Equal Opportunity EmployerThis employer is required to notify all applicants of their rights pursuant to federal employment laws. For further information, please review the Know Your Rights notice from the Department of Labor.#J-18808-Ljbffr