Solutions Architect

AI Full Stack Architect

Make your application after reading the following skill and qualification requirements for this position.

Status: Must be legally eligible to work where you live.

Resume must be in English

Location

Remote, some travel to Atlanta, GA

Engagement

Contract, Long Term, Full Time

Job Description — AI Center of Excellence

About the Role

We are seeking an AI Full Stack Architect to join our AI Center of Excellence (AI COE) — a senior, hands-on technical leader who can both design enterprise-grade AI systems and roll up their sleeves to build them. This is not a whiteboard-only role. You will own the architecture of our agentic AI platform end-to-end — from cloud infrastructure and backend services to LLM pipelines, agent orchestration, and front-end experiences — while setting the technical direction for the team around you.

This role is ideal for someone who has built and shipped production AI agents, has deep AWS Bedrock expertise, and can lead architecture decisions while staying close to the code.

Level & Scope

• Seniority: Architect / Principal Engineer
• Experience: 10+ years in software engineering; 4+ years in AI/ML engineering
• Experience building autonomous bots - not just agents
• Scope: Platform-wide — agents, LLM/SLM integration, observability, full stack
• Leadership: Technical lead; mentors engineers, drives standards and reviews
• Ambiguity: Defines the solution when none exists; navigates fast-moving AI landscape
• Impact: Decisions affect platform reliability, cost, and AI product quality at scale

Key Responsibilities

AI Agent Architecture & Development

• Architect and lead the design of autonomous, multi-step, and tool-using AI agent systems — including agent topology, orchestration patterns, memory strategies, and failure recovery.
• Define and enforce standards for ReAct, plan-and-execute, and multi-agent collaboration patterns across the platform.
• Own hands-on development of complex agent pipelines using AWS Bedrock Agents, LangChain, LangGraph, AutoGen, CrewAI, or equivalent frameworks.
• Design agent guardrails, safety layers, and fallback mechanisms for production reliability.

LLM & SLM Integration

• Lead the evaluation, selection, and integration of Large Language Models (e.g., Claude, GPT-4, Llama, Mistral) and Small Language Models (e.g., Phi-3, Gemma, Mistral 7B) into production workflows.
• Own prompt engineering standards, few-shot learning strategies, RAG (Retrieval-Augmented Generation) pipeline architecture, and fine-tuning workflows.
• Drive model benchmarking and cost/capability trade-off decisions at the platform level.

Cloud & Infrastructure (AWS)

• Architect and own AI workloads on AWS Bedrock — including foundation model access, Knowledge Bases, and Bedrock Agents.
• Design scalable, secure, and cost-efficient cloud infrastructure leveraging Lambda, API Gateway, S3, IAM, ECS/EKS, and related AWS services.
• Define infrastructure-as-code standards and CI/CD pipelines for AI model and agent deployments.

Full Stack Development

• Lead architecture and hands-on development of full stack AI-powered applications — responsive front-end UIs, backend APIs, and agent orchestration layers.
• Set standards for Node.js backend services, Python agent and ML pipelines, and modern front-end frameworks (React, Next.js, or Vue).
• Architect streaming agent response patterns, tool-call UIs, and real-time agent interaction experiences.

Agent Monitoring, Observability & Evaluation

• Own the agent observability strategy — define what gets measured, how, and why.
• Implement and maintain monitoring pipelines using Fiddler AI and comparable platforms (LangSmith, Arize, Weights & Biases, Helicone).
• Define and track key agent performance metrics: accuracy, latency, hallucination rate, tool-call success rate, and cost per inference.
• Establish evaluation frameworks for continuous model and agent quality assurance.

Technical Leadership & Governance

• Lead architecture reviews, define technical standards, and maintain platform documentation.
• Mentor and upskill engineers across the AI COE on agent development, LLM best practices, and responsible AI.
• Collaborate with product, engineering, and business stakeholders to translate complex requirements into actionable, scalable AI solutions.
• Contribute to the organization's responsible AI principles — bias detection, explainability, and model governance.

Required Qualifications

Cloud & Infrastructure

• AWS Bedrock — deep, hands-on experience building and deploying agents and models (foundation model access, Knowledge Bases, Bedrock Agents).
• Strong command of the broader AWS ecosystem: Lambda, S3, IAM, API Gateway, ECS/EKS, CloudWatch.
• Experience designing secure, scalable cloud architectures for AI workloads.

Agent Development

• Proven, hands-on experience architecting and building AI agents — autonomous agents, tool-calling agents, ReAct / plan-and-execute patterns, and multi-agent orchestration systems.
• Deep expertise with agentic frameworks: LangChain, LangGraph, AutoGen, CrewAI, or AWS Bedrock Agents.
• Strong understanding of agent memory, context window management, tool use, and state machines.
• Experience designing agent guardrails, safety constraints, and production-grade failure handling.

Languages & Frameworks

• Python — expert-level; ML pipelines, agent logic, data processing, API development, and scripting.
• Node.js — expert-level; API services, backend orchestration layers, streaming integrations.
• Front-end — working proficiency in React, Next.js, or Vue for AI-powered UIs and agent chat experiences.

LLM / SLM Expertise

• Hands-on production experience with Large Language Models (Claude, GPT-4, Llama, Mistral) and Small Language Models (Phi-3, Gemma, Mistral 7B).
• Deep practical knowledge of prompt engineering, few-shot learning, RAG pipelines, and fine-tuning.
• Experience benchmarking models and making architecture-level model selection decisions.

Monitoring & Observability

• Hands-on experience with Fiddler AI or comparable agent/model monitoring platforms (LangSmith, Arize, Weights & Biases, Helicone).
• Ability to define, implement, and own agent performance metric frameworks at the platform level.

Preferred Qualifications

• Experience with vector databases (Pinecone, pgvector, OpenSearch, Weaviate) for semantic search and RAG architectures.
• Strong MLOps background — CI/CD for models, model versioning, A/B evaluation, drift detection.
• Exposure to multi-modal models (vision + language) and their integration patterns.
• Prior experience leading AI platform or product engineering at scale in a full stack capacity.
• Experience designing streaming response architectures and real-time agent UIs.
• AWS Certifications: AWS Certified Solutions Architect — Professional and/or AWS Certified Machine Learning — Specialty strongly preferred.

What We're Looking For

• Hands-on Architect: You write production code, not just diagrams — you ship agents and own them post-launch.
• Architectural Thinker: You see the system holistically — trade-offs, failure modes, scalability, and cost at once.
• Curiosity-Driven: You track the AI landscape weekly and bring new frameworks and ideas to the team proactively.
• Ownership Mindset: You take a feature from idea to production and care about reliability, cost, and quality long after launch.
• Multiplier: You raise the technical bar of every engineer around you through reviews, mentorship, and documentation.
• Clear Communicator: You translate complex AI behavior and architecture decisions for non-technical stakeholders without losing precision.

Why Join Us

• Architect and shape cutting-edge agentic AI systems deployed at enterprise scale.
• Be a founding technical voice in the AI Center of Excellence — a team dedicated to pushing the boundaries of what AI can do inside and outside the organization.
• Collaborative, low-ego culture with a strong bias for shipping real things.
• Competitive compensation, flexible work arrangements, and a dedicated continuous learning budget. xhyhwjd
• Direct influence over the AI platform roadmap and technology choices.