Gen AI Architect||Onsite

Description

Core GenAI & LLM Skills

• Hands‑on expertise with Amazon Nova / Nova 2 foundation models on Amazon Bedrock, including multimodal and high‑reasoning variants.
• Experience customizing or fine‑tuning LLMs using Bedrock + SageMaker SFT pipelines.
• Strong LLM prompt engineering, optimization, prompt routing, and evaluation skills.

Conversational AI (Chat + Voice)

• Proven ability designing and building conversational agents with Amazon Lex V2 (Assisted NLU, multi‑turn dialogs, analytics).
• Hands‑on integration of Lex with Amazon Connect for IVR/voice bots and contact center automation.
• Experience leveraging Bedrock Knowledge Bases to enable Retrieval‑Augmented Generation (RAG) for grounded responses.

Bedrock Agents / Agentic Systems

• Experience building Amazon Bedrock Agents with ReAct planning or custom orchestration strategies.
• Ability to configure action groups (API schema + Lambda functions) and design multi‑step task workflows.
• Understanding of Bedrock AgentCore for deterministic policy enforcement and safe autonomous execution.

Governance, Safety & Security

• Implementation of Amazon Bedrock Guardrails, including harmful‑content filtering, denied topics, PII redaction, and grounding checks.
• Knowledge of AI governance for regulated industries (banking/finance), including safety policies and audit controls.
• Strong IAM, encryption, data isolation, and compliance patterns for enterprise AI.

RAG, Data & Integrations

• Building RAG systems using Bedrock Knowledge Bases (document ingestion, chunking, embedding.
• Integrations using AWS Lambda, API Gateway, DynamoDB, S3, and event‑driven patterns to support conversational flows.
• Experience optimizing LLM latency and cost via model selection, caching, and cross‑region inference.
  Experience with Connect + Lex + Bedrock Agent architectures for real‑time voice assistants.

• Understanding of near‑real‑time conversational patterns informed by Nova‑powered speech capabilities.

Operational Excellence (MLOps / LLMOps)

• Strong skills in LLM evaluation, monitoring, telemetry, and continuous quality measurement.
• Experience with Lex Analytics dashboards for intent/slot performance.
• Ability to version prompts, guardrails, agent instructions, and model configs.

Optional Search Keywords (Good-to-Have)

• “Bedrock Agents custom orchestration”
• “Lex V2 Assisted NLU / QnAIntent generative AI”
• “Nova 2 reasoning model experience”
• “Amazon Connect AI integration”
• “AgentCore policy enforcement / governance”
• “Multi-agent orchestration frameworks”