AI Solution Architect

The AI Solution Architect works side by side with talented Product, Design and Engineering teams to deliver AI‑driven solutions that delight customers and accelerate delivery, balancing feasibility, scalability, and compliance. Guide engineering teams on when and how to apply AI tools—copilots, agents, and vendor integrations—to real business problems while enforcing architectural standards. The AI Solution Architect translates strategy into architecture, patterns, and working software that deliver measurable outcomes. Works with code, coach engineers and tackle obstacles to success. Key Responsibilities: Own end‑to‑end solution architectures for AI and AI‑enabled products (discovery → design → deployment), ensuring security, reliability, cost efficiency, and maintainability across cloud and on‑prem boundaries. Establish reference architectures and reusable patterns for GenAI and applied AI (RAG, agents/orchestration, vector search, prompt & tool design, event‑driven microservices, API gateways). Select fit‑for‑purpose models and services (e.g., Azure OpenAI/Bedrock/Vertex, OSS LLMs, embedding models) with clear tradeoffs on performance, latency, privacy, and cost. Partner with product and platform teams to ship solutions: define requirements, review designs and PRs, and drive prototype → pilot → production with CI/CD, IaC, and MLOps/LLMOps (model versioning, prompt/config management, evals, drift & safety monitoring). Coach teams to use copilots (e.g., GitHub/Claude), agent frameworks (e.g., LangGraph/Semantic Kernel), and integration SDKs responsibly to improve velocity without compromising quality or security. Ensure observability (tracing, guardrails, red‑teaming, cost dashboards), SLOs, and runbooks are in place before cutover. Run architecture discovery with business stakeholders; frame problems, quantify constraints, and map KPIs (time‑to‑first‑value, cost‑to‑serve, task success, CSAT/NPS, deflection rate, accuracy). Communicate tradeoffs and roadmaps in clear language to executives and non‑technical partners; publish decision records and architecture docs. Raise the bar on engineering excellence—coding standards, design reviews, threat modeling, and documentation. Active coach and mentor to engineers, evangelizing a culture of pragmatic, responsible innovation; contribute to communities of practice. Qualifications Education & Experience: Bachelor’s degree in Computer Science, Engineering, Data Science, Artificial Intelligence, or equivalent experience. 8+ years in software/solution architecture or platform engineering, including 3+ years delivering applied ML/GenAI solutions in production environments. Extensive hands-on experience with Google Cloud Platform (GCP), including Vertex AI, BigQuery, Dataflow, Pub/Sub, and cloud-native microservices, APIs, event streaming, containers/orchestration (Kubernetes/GKE), and Infrastructure as Code (Terraform/Deployment Manager). Practical expertise with GenAI patterns: Retrieval-Augmented Generation (RAG), vector databases (e.g., BigQuery Vector Search), prompt engineering/evaluation, agent design, function/tool calling, and orchestration using Google AI tools. Strong grasp of MLOps/LLMOps: CI/CD for models and prompts, offline/online evaluation, telemetry, drift and safety monitoring, with a focus on Google’s Vertex AI Pipelines, Model Registry, and continuous deployment. Skills & Abilities: Experience in regulated industries (financial services, healthcare), contact center/knowledge management, or workflow automation. Certifications (nice‑to‑have): GCP Architect, Security (e.g., CCSK) or equivalent Experience designing for security, privacy, and compliance within GCP; proficiency with OAuth/OIDC, secrets management (Secret Manager), data protection, and model/content safety controls aligned with Google’s best practices. Exceptional written and verbal communication skills; proven ability to influence and collaborate across product, engineering, security, and business teams. Other Requirements: Ability to work occasional overtime. Occasional travel (up to ~15%). Occasional after‑hours work to support releases or incident response. Prolonged periods of sitting at a desk and working on a computer.