We are seeking an exceptional Senior AI Engineer with 6-8 years of experience to architect and deliver enterprise-grade AI systems at scale, with primary focus on Amazon Bedrock and Microsoft Azure AI Foundry -- the two leading managed cloud AI platforms. This is a staff-level individual contributor position for engineers who can drive technical strategy, mentor cross-functional teams, and translate the latest advances in generative AI into high-impact production systems.
You will operate as a principal builder and technical authority across our most ambitious AI initiatives: multi-agent orchestration pipelines, RAG-powered knowledge systems, LLM fine-tuning workflows, and cloud-native AI infrastructure. You will own decisions that reach millions of users and set the engineering benchmark for AI excellence across the organization.
This role intentionally avoids SageMaker-centric workflows in favor of the latest serverless and managed AI services -- Amazon Bedrock AgentCore, Amazon Nova, Amazon S3 Vectors, Microsoft Foundry, and Azure AI Agent Service -- enabling faster iteration, lower infrastructure overhead, and access to frontier models.
Technical Leadership & AI Architecture
- Own end-to-end architecture of cloud-native AI systems on AWS and Azure, from raw data ingestion through LLM serving, evaluation, and continuous monitoring.
- Establish AI engineering standards, model governance frameworks, and cloud design patterns adopted across multiple product teams.
- Lead critical cross-platform design decisions: when to use AWS Bedrock vs. Azure AI Foundry, which foundation model best fits performance and cost constraints, and when to fine-tune vs. prompt-engineer.
- Evaluate and rapidly prototype with emerging AWS and Azure AI capabilities (Amazon Nova 2, Bedrock Reinforcement Fine-Tuning, Azure GPT-5-Codex, Foundry Agent Service) to keep the organization at the frontier.
- Drive technical design reviews, architecture critiques, and risk assessments for high-stakes AI deployments.
AWS Amazon Bedrock & Generative AI Engineering
- Architect enterprise applications using Amazon Bedrock as the primary LLM serving and orchestration platform -- leveraging the full model catalog including Amazon Nova 2 (Lite, Pro, Sonic, Omni), Anthropic Claude, Mistral Large 3, and Amazon Bedrock Marketplace models.
- Build production multi-agent systems using Amazon Bedrock AgentCore: design agent boundaries with Policy controls, implement episodic Memory for long-horizon reasoning, and configure AgentCore Gateway for secure tool integrations with Salesforce, Slack, and internal APIs.
- Implement browser-based workflow automation agents with Amazon Nova Act, achieving 90%+ task reliability for form filling, data extraction, QA testing, and enterprise UI automation workflows.
- Design and deploy RAG pipelines using Amazon Bedrock Knowledge Bases with Amazon S3 Vectors as the primary vector store -- supporting up to 2 billion vectors per index with 100ms query latencies at up to 90% lower cost than specialized vector databases.
- Apply Amazon Bedrock Reinforcement Fine-Tuning for domain-specific model customization, achieving up to 66% accuracy gains over base models without requiring labeled datasets or deep ML expertise.
- Leverage Amazon Bedrock Model Distillation to create task-specific models that run up to 500% faster and cost 75% less, with minimal accuracy trade-off.
- Implement Amazon Bedrock Intelligent Prompt Routing and Prompt Caching to reduce inference costs by up to 30% while maintaining response quality.
- Apply Amazon Bedrock Guardrails for responsible AI enforcement: content filtering, hallucination prevention, PII redaction, and compliance controls across all production LLM endpoints.
- Develop and manage Kiro-based agentic development workflows: context-aware, session-persistent autonomous agents for coding and operational task automation.
AZURE Microsoft Foundry & Azure AI Engineering
- Build and deploy intelligent applications on Microsoft Foundry (formerly Azure AI Studio), the unified platform for model access, fine-tuning, evaluation, and agent deployment at enterprise scale.
- Integrate and orchestrate frontier models through Foundry Models: Azure OpenAI GPT-4.1 (1M token context), o3 / o4-mini reasoning models, GPT-5-Codex for multimodal code reasoning, Anthropic Claude (Sonnet 4.5, Opus 4.1), and Mistral Large 3.
- Design and operate multi-agent systems using Microsoft Agent Framework (Semantic Kernel + AutoGen runtime), building stateful long-running agents deployed via Hosted Agents and published to Microsoft 365 Copilot with one-click.
- Build agentic workflows with Azure AI Agent Service, connecting agents to enterprise data via Azure AI Search (RAG), line-of-business APIs through Azure API Management, and real-time data through Azure Event Hubs.
- Leverage Microsoft Foundry Control Plane for fleet-wide agent observability, governance, policy enforcement, and compliance auditing across all deployed AI agents.
- Implement Deep Research agents in Microsoft Foundry Agent Service as composable, programmable research engines embedded in multi-agent workflows and enterprise applications.
- Use Foundry Tools (unified suite of prebuilt AI capabilities for audio, video, image, document, and text) to accelerate development of intelligent agents without rebuilding common AI tasks.
- Apply Azure AI Foundry Model Router to dynamically route prompts to the optimal model (GPT-4.1, o4-mini, Phi-4) at runtime, minimizing costs while preserving quality SLAs.
- Configure NSP-protected private deployments using VNETs, private endpoints, and BYO Key Vault connections for regulated-industry AI workloads on Foundry.
Cross-Platform Infrastructure & MLOps
- Architect AWS-native training infrastructure using EC2 Trainium3 UltraServers (Trn3) -- AWS's 3nm AI chip delivering 4x performance over Trn2 -- for large-scale model training and inference at lower cost.
- Build and manage containerized model serving on Amazon EKS with Karpenter autoscaling, integrated with Triton Inference Server for multi-model, multi-framework inference optimization.
- Implement fully serverless inference using AWS Lambda Managed Instances for on-demand LLM endpoint invocations without infrastructure provisioning.
- Establish IaC-driven ML platform provisioning using AWS CDK (Python) and Azure Bicep/Terraform, ensuring all AI environments are version-controlled, reproducible, and auditable.
- Design cross-cloud observability stacks: Amazon CloudWatch + AWS X-Ray on AWS Azure Monitor + Application Insights on Azure -- with unified dashboards for model drift detection, latency SLOs, and cost alerting.
- Implement MLflow (serverless via SageMaker AI integration or self-hosted on EKS) for experiment tracking, model versioning, and lifecycle management across both AWS and Azure workloads.
- Champion cloud cost optimization: AWS Spot instance strategies and Bedrock Intelligent Prompt Routing for AWS workloads Azure Developer Training tier and Serverless API pay-as-you-go for Azure workloads.
Evaluation, Safety & Responsible AI
- Define and own the organization's AI evaluation framework -- automated pipelines measuring model accuracy, fairness, bias, toxicity, hallucination rate, and safety across both AWS Bedrock and Azure Foundry deployments.
- Lead red-teaming, adversarial robustness testing, and jailbreak analysis using Amazon Bedrock Guardrails and Azure AI Content Safety before every production launch.
- Establish model risk management: model cards, prompt audit trails, production monitoring dashboards, and incident response runbooks for AI systems.
- Collaborate with Legal, Compliance, and the AI Ethics board to meet regulatory requirements (EU AI Act, NIST AI RMF, ISO 42001, SOC 2 Type II).
Mentorship & Organizational Influence
- Serve as principal technical mentor for mid and junior AI engineers: conducting code reviews, architecture design sessions, and cloud platform enablement workshops.
- Represent the team at executive forums, customer briefings, and external AI conferences.
- Partner with Product, Research, Data Science, and Platform Engineering to align AI roadmap with business objectives and cloud investment strategy.
- Drive org-wide knowledge sharing via internal tech talks, open-source contributions, technical blog posts, and AWS/Azure certification mentorship programs.
- Lead technical hiring: define role requirements, design AI cloud-focused interview loops, and make recommendations for senior and staff-level candidates.
Our Interview Practices
