Solutions Architect - AI Platform

Duties: Own the end-to-end architecture of the enterprise AI platform on Azure: AI gateway and model routing, generative AI and machine learning services, agent orchestration, memory and retrieval, data integration, guardrails, security, evaluation and observability. Define the minimum platform needed to start building the first workflows, then lay out the roadmap for later capabilities such as a semantic layer, centralized orchestration and an evaluation framework. Set the standards every workflow team follows: how AI solutions and agents are onboarded, how autonomy and human review are set per workflow, how tenancy and data isolation work, and how models and agents are evaluated. Design integrations with our core fund administration systems, Snowflake data platform, MCP servers, document stores and third-party APIs. Partner with security, compliance and risk teams so the platform meets our regulatory, data-protection and audit requirements from day one. Review designs from the AI Development pods and make build-versus-buy and vendor recommendations. Work alongside external partners during the initial build and make sure knowledge transfers to the Ultimus team.

Requirements: 8+ years in software or cloud architecture, including 2+ years designing production AI, ML or LLM-based systems. •Deep hands-on experience with Microsoft Azure (AKS, API Management, networking, identity and Key Vault). •Working knowledge of modern AI application patterns: machine learning operations, retrieval-augmented generation, agent orchestration, tool calling, prompt and context management, and evaluation. •Experience designing secure, multi-tenant platforms that handle sensitive financial or personal data. •Strong data architecture background, including data warehouses and integration patterns. • Ability to explain technical trade-offs clearly to engineers, operations leaders and executives.