Agentic AI Solution Architect

Agentic AI Solution Architect (AWS Bedrock | MultiModel | AIDLC/VibeCoding)

Atlanta, GA (Onsite 3 days)

Role Summary

We are looking for a hands-on Agentic AI Solution Architect to design and lead implementation of a multi-model agentic AI framework on AWS Amazon Bedrock, enabling real-time, governed, secure execution of AI agents through an orchestrator-led agent coordination layer. This role will define architecture patterns, security/guardrails, AIDLC delivery approach, and production readiness for next-gen AI applications.

Key Responsibilities

• Own end-to-end architecture for multi-agent, multi-model systems, including agent orchestration, routing, and execution flows.
• Design and implement orchestrator-based agent coordination, ensuring consistent execution across core agents (e.g., decisioning flows such as fraud/funding/capital) and supporting agents.
• Define Model Abstraction & Governance Layer using AWS Amazon Bedrock (model routing, guardrails, knowledge bases/RAG integration where applicable).
• Establish AI safety and security guardrails (e.g., injection protections, PII masking, policy/guardrail compliance) and ensure they are enforced at design + runtime.
• Lead AIDLC (AI Development Life Cycle) adoption: vibe coding agentic coding AI-CI AI-CD monitoring, with human-in-the-loop quality gates.
• Architect testing strategy using testing agents: unit, integration, performance and security validations; ensure shift-left production readiness before deployment.
• Define non-functional requirements: latency/SLAs, scalability, reliability, and observability, including drift/hallucination detection and adaptive feedback in production.
• Collaborate with product, engineering, QA, and stakeholders to convert business workflows into robust agentic solutions (including explainable decision logic).

Required Skills & Experience

• Proven experience architecting AI/GenAI solutions with a strong understanding of agentic systems and orchestration patterns.
• Strong hands-on with AWS Amazon Bedrock concepts (model execution, guardrails/governance, scalable inference patterns).
• Experience implementing secure AI systems: PII handling, injection risk mitigation, policy enforcement, access controls.
• Strong engineering background in modern backend architecture and APIs (service patterns, orchestration). (Framework examples in deck include FastAPI/LangGraph-style orchestration patterns.)
• Ability to define and enforce testing & validation frameworks for AI agent ecosystems.
• Experience with multi-agent collaboration approaches for code generation/refactoring (agentic coding) and vibe coding-style scaffolding.

Good-to-Have

• Fleet management domain / functional background will be PLUS
• Credit cards / Banking products knowledge will be PLUS
• Experience with real-time systems, event-driven workflows, or systems with strict response-time expectations.