AI Full Stack Architect Agentic AI & AWS Bedrock

JKV International LLC is seeking an experienced AI Full Stack Architect to lead the architecture and development of enterprise-grade Generative AI applications, autonomous AI agents, and intelligent software platforms.

This is a senior, hands-on architecture role for a technical leader who can design scalable AI solutions, build production-ready agentic systems, and establish architectural standards across cloud infrastructure, backend services, LLM integration, agent orchestration, and frontend applications.

The ideal candidate brings deep, practical experience with AWS Bedrock, agentic AI architectures, Python, Node.js, RAG pipelines, and enterprise cloud solutions. You will be responsible for translating complex business requirements into robust technical architectures, making critical technology decisions, guiding engineering teams, and ensuring that AI solutions meet production standards for reliability, security, performance, and cost efficiency., AI Architecture & Agentic AI Engineering

• Architect and implement enterprise-grade AI agents, autonomous workflows, tool-calling systems, and multi-agent orchestration platforms.
• Define agent architecture patterns, including ReAct, plan-and-execute, agent memory, state management, context handling, and failure recovery.
• Design reusable agent frameworks, orchestration layers, integration patterns, and enterprise AI components.
• Implement production-grade guardrails, human-in-the-loop workflows, fallback mechanisms, and error-handling strategies.
• Evaluate architectural trade-offs involving scalability, reliability, security, performance, and operational cost.

AWS Cloud Architecture & Infrastructure

• Architect, deploy, and manage AI solutions using AWS Bedrock, foundation models, Knowledge Bases, and Bedrock Agents.
• Design secure, scalable cloud-native solutions using AWS Lambda, API Gateway, S3, IAM, ECS/EKS, CloudWatch, and related AWS services.
• Define infrastructure architecture, deployment strategies, environment management, and infrastructure-as-code standards.
• Establish CI/CD pipelines and deployment practices for AI applications, agent workflows, and backend services.
• Optimize cloud infrastructure for availability, latency, throughput, security, and cost efficiency.

LLM, SLM & Generative AI Solutions

• Evaluate and integrate Large Language Models and Small Language Models into production applications.
• Define standards for prompt engineering, structured outputs, tool calling, context management, and model integration.
• Architect Retrieval-Augmented Generation (RAG) pipelines, embeddings, semantic search, and enterprise knowledge retrieval systems.
• Lead model benchmarking and evaluate trade-offs between accuracy, latency, capabilities, scalability, and cost.
• Assess fine-tuning, model adaptation, and multimodal AI approaches where appropriate.

Full Stack Application Development

• Lead the architecture and implementation of end-to-end AI-powered applications spanning frontend interfaces, backend APIs, and AI orchestration services.
• Establish backend engineering standards using Python and Node.js.
• Guide modern frontend architecture using React, Next.js, Vue, or equivalent frameworks.
• Design streaming AI responses, real-time interactions, tool-call interfaces, and scalable application integrations.
• Ensure application security, maintainability, testability, performance, and production readiness.

AI Observability, Evaluation & Governance

• Define observability and evaluation strategies for AI models, agent execution, and production applications.
• Implement or guide monitoring solutions using Fiddler AI or comparable platforms such as LangSmith, Arize, Weights & Biases, or Helicone.
• Establish metrics for response quality, hallucination rates, latency, tool-call success, failure rates, token consumption, and cost per inference.
• Build evaluation and regression-testing frameworks for continuous quality improvement.
• Implement monitoring, alerting, responsible AI controls, and model governance practices.

Technical Leadership & Architecture Governance

• Own end-to-end architecture decisions and drive technology selection for enterprise AI solutions.
• Lead architecture reviews, technical design discussions, code reviews, and engineering standards.
• Develop reference architectures, reusable design patterns, technical documentation, and best practices.
• Mentor senior engineers and development teams on AI architecture, agentic systems, cloud engineering, and responsible AI.
• Collaborate with business, product, security, and engineering stakeholders to deliver scalable, maintainable AI solutions., * Own AI architecture decisions from initial design through production deployment and ongoing optimization.
• Explain and defend architecture decisions involving AWS Bedrock, agent orchestration, model selection, RAG, and cloud infrastructure.
• Troubleshoot real-world challenges involving hallucinations, context limits, latency, tool failures, security, and operating costs.
• Build scalable AI platforms with measurable improvements in performance, reliability, maintainability, and business outcomes.
• Set technical direction, challenge design assumptions, mentor engineers, and raise engineering quality.
• Balance strategic architecture responsibilities with hands-on implementation and production ownership.

• 10+ years of professional software engineering experience, including substantial system design and architecture responsibilities.
• 4+ years of hands-on AI/ML or Generative AI engineering experience, with demonstrable production implementations.
• Experience operating as an Architect, Principal Engineer, Lead Architect, or equivalent senior technical leader.
• Deep, hands-on AWS Bedrock experience, including foundation model integration, Knowledge Bases, and/or Bedrock Agents.
• Proven experience designing and implementing production-grade AI agents, tool-calling workflows, or multi-agent systems.
• Strong expertise with agentic AI frameworks such as LangGraph, LangChain, AutoGen, CrewAI, or AWS Bedrock Agents.
• Expert-level Python skills for AI services, agent logic, backend APIs, integrations, and data processing.
• Strong Node.js expertise for backend services, API development, and orchestration layers.
• Practical experience with React, Next.js, Vue, or comparable frontend technologies.
• Strong understanding of LLM integration, prompt engineering, RAG, embeddings, semantic retrieval, and model evaluation.
• Solid knowledge of AWS cloud architecture, IAM, Lambda, API Gateway, S3, and relevant container and monitoring services.
• Experience designing secure, scalable, resilient, and cost-efficient enterprise applications.
• Experience with AI observability, model evaluation, production monitoring, or comparable quality assurance frameworks.
• Ability to communicate architectural decisions, technical trade-offs, and implementation strategies to technical and business stakeholders.
• Demonstrated ability to lead engineering delivery while remaining hands-on with architecture, coding, design reviews, and troubleshooting., * Experience with vector databases and search technologies such as Amazon OpenSearch, Pinecone, pgvector, or Weaviate.
• Experience with MLOps/LLMOps, automated deployments, model versioning, evaluation pipelines, and performance monitoring.
• Experience with multimodal AI, advanced retrieval strategies, and enterprise knowledge systems.
• Knowledge of Terraform, AWS CDK, CloudFormation, or other infrastructure-as-code technologies.
• Experience with enterprise security, data privacy, identity management, responsible AI, and governance.
• Previous experience architecting AI platforms or products supporting multiple teams or enterprise-scale workloads.
• AWS Solutions Architect Professional and/or AWS Machine Learning certifications are advantageous., Candidates should have demonstrable experience designing and delivering production AI solutions, owning architecture decisions, and leading complex engineering initiatives. Experience limited to academic projects, basic chatbot development, tutorials, or proof-of-concept implementations without substantial architecture ownership will not meet the expectations of this role.