AI Engineer - Salem - Cloud202

Description

Location: Remote | Job Type: Full-Time | Experience Level: 3+ years 

About Us 

Cloud202 Limited is a leading technology consulting company dedicated to helping businesses transform and innovate through cutting-edge technology solutions. We specialize in cloud migration, AI/ML, and application development, providing our clients with the expertise they need to stay ahead in a rapidly evolving digital landscape. 

Position Overview 

We are seeking an innovative AI Engineer to lead the development and implementation of enterprise-grade agentic AI solutions. This role requires deep expertise in the Gen-AI ecosystem, including Amazon Bedrock, Amazon Bedrock AgentCore, SageMaker AI, and emerging AI agent frameworks. The ideal candidate will drive enterprise AI transformation initiatives and build next-generation intelligent applications using cutting-edge agentic platforms and protocols. 

Required Qualifications 

Experience 

• Minimum 3+ years of hands-on experience with AWS cloud services and machine learning infrastructure 
• 2+ years of specific experience with generative AI, large language models (LLMs), and foundation models 
• Proven track record of building and deploying production-scale AI/ML applications on AWS 

Certifications 

• Preferred: AWS Certified AI Practitioner or AWS Machine Learning Specialty 

Core Technical Skills 

Amazon Bedrock AgentCore Platform (Critical) 

• AgentCore Runtime: Deploy and operate AI agents securely at scale with serverless infrastructure, session isolation, and support for 8-hour execution windows 
• AgentCore Memory: Implement intelligent session and long-term memory with episodic learning capabilities for context-aware agent interactions 
• AgentCore Gateway: Build secure, centralized access to tools and APIs with minimal code transformation 
• AgentCore Identity: Implement seamless agent authentication across AWS services and third-party applications (Slack, Zoom, GitHub, Salesforce) using OAuth, Okta, Entra, or Amazon Cognito 
• AgentCore Tools: Utilize Code Interpreter for secure code execution and Browser Tool for enterprise-grade web automation within managed sandbox environments 
• AgentCore Observability: Implement end-to-end tracing, debugging, and monitoring through unified CloudWatch dashboards with OTEL compatibility 
• AgentCore Policy: Set fine-grained boundaries on agent actions with real-time deterministic controls 
• AgentCore Evaluations: Continuously assess agent quality and behavior for production readiness 

Gen-AI Services & Foundation Models 

• Amazon Bedrock: Comprehensive experience with foundation model access, fine-tuning, and deployment 
• SageMaker AI: Model hosting, endpoints, auto-scaling, A/B testing, and deployment pipelines 
• Amazon Q Developer: AI-powered development automation and code transformation capabilities 
• Foundation Models: Hands-on experience with Claude (Anthropic), Llama (Meta), GPT models (OpenAI), Mistral, and Amazon Nova models 

AI Agents Development & Frameworks 

• Strands Agents SDK: Build production-ready AI agents with model-driven approach, supporting single agents, multi-agent systems, and swarm architectures 
• Framework Expertise: Experience with CrewAI, LangGraph, LlamaIndex, Google ADK, OpenAI Agents SDK, or custom agent frameworks 
• Multi-Agent Orchestration: Design complex workflows with hierarchical delegation, agent-as-tools patterns, and dynamic capability discovery 
• Agentic Workflows: Build autonomous agents that reason, plan, use tools, and maintain context across long-running tasks 
• Tool Integration: Develop custom tools using Python decorators and integrate external APIs and services 

Agent Protocols & Interoperability (Essential) 

• Model Context Protocol (MCP): Implement MCP servers and clients to provide standardized context and tool access to AI agents. Deploy MCP servers in AgentCore Runtime with OAuth authentication 
• Agent-to-Agent (A2A) Protocol: Build inter-agent communication systems using A2A protocol for peer-to-peer agent collaboration, capability negotiation, and task coordination 
• Agent Discovery: Implement agent cards and capability manifests for dynamic agent discovery and routing 
• Protocol Integration: Deploy agents supporting both MCP and A2A protocols for maximum interoperability across enterprise systems 

Advanced Technical Skills 

• Vector Databases: Amazon OpenSearch, Pinecone, or similar for RAG implementations 
• Programming: Expert-level Python and JavaScript/TypeScript, with focus on AI/ML libraries and async programming 
• APIs & Integration: RESTful APIs, GraphQL, JSON-RPC 2.0, Server-Sent Events (SSE), real-time streaming, webhook integration 
• Prompt Engineering: Advanced prompt flows, few-shot learning, chain-of-thought reasoning, and structured output generation 
• Knowledge Bases: RAG implementation with enterprise data integration and semantic search 
• Guardrails & Safety: Bedrock Guardrails, content filtering, bias detection, and responsible AI practices 
• Custom Model Fine-tuning: Adapting foundation models for domain-specific use cases 

Advanced GenAI Applications 

• Retrieval-Augmented Generation (RAG): Enterprise search, document Q&A, knowledge management 
• Content Generation: Text, image, code, and multimedia content creation 
• Conversational AI: Chatbots, virtual assistants, customer service automation with memory retention 
• Code Generation & Analysis: Automated code review, documentation, refactoring, and software modernization 
• Data Analysis & Insights: Natural language to SQL, automated reporting, business intelligence 

Key Responsibilities 

Solution Architecture & Design 

• Design end-to-end generative AI solutions using Amazon Bedrock AgentCore as the primary agentic platform 
• Architect scalable, cost-effective AI pipelines leveraging AgentCore Runtime for serverless deployment 
• Implement MCP and A2A protocols for agent interoperability and tool integration 
• Design multi-agent architectures with proper orchestration, memory management, and observability 
• Create technical documentation and best practices for AgentCore implementations 

Development & Implementation 

• Build production-ready agentic applications using Amazon Bedrock AgentCore services (Runtime, Memory, Gateway, Identity, Observability) 
• Develop AI agents using Strands Agents SDK and other framework-agnostic approaches 
• Implement MCP servers for tool and data access across enterprise systems 
• Deploy A2A-compliant agents for cross-platform agent collaboration 
• Implement RAG systems with vector databases and AgentCore Gateway for secure data access 
• Create automated workflows for model deployment, monitoring, and evaluation 
• Integrate AI capabilities into existing enterprise applications with proper authentication and governance 

Model & Agent Management 

• Evaluate and select appropriate foundation models for specific use cases 
• Implement AgentCore Policy for fine-grained control over agent actions and permissions 
• Use AgentCore Evaluations for continuous quality assessment and optimization 
• Optimize agent performance, cost, and latency using AgentCore Observability insights 
• Ensure compliance with data privacy, security requirements, and responsible AI practices 

Innovation & Research 

• Stay current with latest AWS AI service releases, AgentCore capabilities, and agentic AI protocols 
• Experiment with emerging AI techniques, multi-agent patterns, and protocol enhancements 
• Prototype new use cases and proof-of-concepts using AgentCore platform 
• Contribute to internal AI strategy, AgentCore best practices, and community open-source projects 

Preferred Experience & Skills 

Industry-Specific Knowledge 

• Experience with industry-specific AI applications (healthcare, finance, retail, manufacturing) 
• Understanding of compliance requirements (GDPR, HIPAA, SOX, PCI-DSS) 
• Knowledge of AI ethics, bias mitigation, and responsible AI governance 

Advanced Technical Skills 

• MLOps: Model lifecycle management, automated retraining, drift detection with SageMaker Pipelines 
• Real-time AI: Streaming data processing, low-latency inference, event-driven architectures 
• Multimodal AI: Text, image, audio, and video processing with Amazon Nova models 
• Edge AI: Model optimization for edge deployment 
• Custom Training: Fine-tuning foundation models with proprietary data 
• Infrastructure as Code: CloudFormation, AWS CDK, or Terraform for AgentCore deployments 

Leadership & Collaboration 

• Experience leading AI transformation initiatives and AgentCore adoption 
• Ability to communicate complex agentic AI concepts to non-technical stakeholders 
• Cross-functional collaboration with product, engineering, and business teams 
• Mentoring junior engineers and data scientists on AgentCore best practices 

Recent Technology Awareness (2025) 

• Amazon Bedrock AgentCore GA release with VPC support, A2A protocol, and enhanced observability (October 2025) 
• Strands Agents SDK 1.0 with multi-agent orchestration, session management, and A2A support 
• Agent-to-Agent (A2A) protocol under Linux Foundation governance with enterprise adoption 
• Model Context Protocol (MCP) enhancements for agent-to-agent communication and tool integration 
• Latest Amazon Nova models (Nova Premier, Nova Sonic) for multimodal and conversational AI 
• Latest Anthropic Claude models (Claude 4 Sonnet) with extended context and enhanced capabilities 
• AgentCore Policy and Evaluations for production-grade agent governance 
• AWS Q Developer CLI integration with MCP for agentic development workflows 

Education & Background 

• Bachelor's degree in Computer Science, AI/ML, Mathematics, or related field 
• Continuous learning mindset with active participation in AI communities and open-source contributions 
• Strong understanding of distributed systems, microservices, and serverless architectures 

Industry: IT Services and IT Consulting 

Employment Type: Full-time