Location: Remote within the USA, or onsite in Buffalo, NY / Wilmington, DE (client preference for candidates near these areas). New hires are required to work onsite at the client's office for the first 2–3 weeks (treated as a business trip; travel expenses covered by the company).
Company Overview Glint Tech Solutions is a women-owned, global IT staffing and recruiting firm serving enterprise clients across the USA and Canada.
Project Description A leading financial services client is seeking a GenAI Solutions Architect to serve as the hands-on consulting architect ensuring AI use cases across the bank are designed and connected the right way — aligned to the enterprise AI platform architecture, approved integration patterns, and governance standards. As business and technology teams stand up AI-enabled applications, this role is their design partner, translating platform capabilities and standards into concrete, buildable solution architectures and reviewing designs before they harden. This is a deeply technical role for an architect who still builds: producing reference architectures, integration patterns, and working examples, while pairing with delivery teams to accelerate adoption and prevent rework and governance escapes.
Key Responsibilities
Serve as the consulting architect for AI use-case teams across the bank, shaping solution designs, data flows, model access patterns, and integration approaches
Produce and maintain reference architectures, design patterns, and working examples for common use-case shapes (RAG applications, document processing, workflow automation, agent-based patterns)
Review solution designs against platform standards and governance requirements before build; document findings and drive remediation with delivery teams
Advise on model selection, prompt/context architecture, retrieval design, and oversight/guardrail patterns appropriate to each use case's risk tier
Ensure all designs route model access through the governed enterprise gateway with correct entitlements, quotas, and logging; prevent parallel or ungoverned access paths
Translate governance standards into architecture requirements delivery teams can implement
Partner with platform engineering on the evolution of platform capabilities based on real use-case demand
Document network, identity, data-classification, and environment-separation considerations for solution designs
Partner with Cybersecurity architecture on AI threat modeling, prompt-injection risk, and adversarial-testing requirements
Pair with application teams that lack AI delivery experience, providing hands-on design and build guidance
Create and deliver architecture enablement materials, design guides, and pattern documentation
Support solution reviews in governance forums with clear, evidence-based architecture assessments
Transfer patterns, documentation, and working knowledge to bank FTEs throughout the engagement
Mandatory Skills
8+ years of experience in solution architecture, application architecture, or senior engineering roles, including hands-on delivery of cloud-native applications
Hands-on experience architecting and delivering GenAI/LLM-based solutions — model integration, RAG pipelines, prompt/context engineering, and agent or workflow patterns
Strong Azure experience — Azure OpenAI/AI services, API Management, Entra ID, Key Vault, networking and landing-zone concepts, environment separation
Demonstrated experience producing reference architectures, integration patterns, and design documentation adopted by multiple teams
Experience designing within security, risk, and compliance constraints in a regulated environment
Strong consulting skills — stakeholder communication, design facilitation, and the ability to influence without authority
Nice-to-Have Skills
Financial services experience and familiarity with banking SDLC governance (design gates, architecture review, permit-to-build/operate models)
Experience with AI gateway/governance patterns — model allowlisting, entitlement-based access, usage controls, audit logging
Experience with vector stores, retrieval services, evaluation harnesses, and MCP/tool-integration patterns