Is the AWS Bedrock Academy free?

Yes. The full course is free and includes 25 modules, 52 lessons, 52 labs, and quiz-based progression with no sign-up required.

How is Infinity AWS Bedrock Academy different from AWS Skill Builder?

Unlike AWS Skill Builder which often gates advanced hands-on labs behind paid subscriptions or AWS Builder ID logins, Infinity Tech Stack's AWS Bedrock Academy is 100% free, ungated, and browser-first with zero sign-up required. It includes 25 modules, 52 lessons, 52 hands-on labs, and 125+ quiz checks covering the Converse API, Bedrock Guardrails, AgentCore, and multi-model evaluations without requiring active AWS account credentials.

What is Amazon Bedrock and how does its architecture operate?

Amazon Bedrock is AWS's fully managed foundation model service providing serverless inference across top providers (Anthropic, Meta, Mistral, Amazon Nova/Titan) via a unified Converse API, backed by AWS IAM, KMS encryption, and VPC PrivateLink isolation.

What is the Amazon Bedrock Converse API and why should you use it over InvokeModel?

The Converse API provides a unified request and response specification across all foundation models on Bedrock. It standardizes system prompts, structured tool use, and multi-turn message history, eliminating vendor-specific payload differences.

How does Amazon Bedrock AgentCore differ from Classic Bedrock Agents?

Classic Bedrock Agents entered maintenance mode (closed to new customers on July 30, 2026), succeeded by Amazon Bedrock AgentCore. AgentCore is AWS's modern, modular agent platform featuring serverless runtimes with sub-second cold starts, zero-boilerplate MCP Gateway tool integration, cross-session episodic memory, and the AgentCore Consent Portal.

How do Bedrock Guardrails enforce safety and PII redaction?

Bedrock Guardrails apply customizable safety policies across both user inputs and model outputs. They block toxic content, enforce Automated Reasoning hallucination checks, and anonymize sensitive PII (emails, SSNs, credit cards) using AWS-managed entity detectors.

What retrieval strategies are supported in Bedrock Knowledge Bases?

Bedrock Knowledge Bases support dense vector retrieval using Titan or Cohere embeddings, BM25 keyword lexical search, native HYBRID search combining vector and sparse rankings, and cross-encoder neural reranking.

How does Bedrock Cross-Region Inference prevent 429 throttling exceptions?

Cross-region inference profiles automatically distribute traffic across multiple regional AWS data centers (e.g., us-east-1, us-west-2, us-east-2), pooling quotas dynamically to absorb traffic spikes without requiring custom client retry code.

What are Amazon Nova models on Bedrock?

Amazon Nova is AWS's flagship model family, including Nova Micro (ultra-fast text), Nova Lite (cost-effective multimodal), Nova Pro (balanced intelligence), and Nova Premier (frontier reasoning), engineered for high-throughput enterprise workloads.

How does prompt caching operate on Amazon Bedrock?

Prompt caching stores frequent prefix tokens in memory for subsequent requests. Cache hits reduce input token billing by up to 90% and lower invocation latency by over 80% on large system prompts, codebases, and legal documents.

What is Provisioned Throughput (PT) and when should it be purchased?

Provisioned Throughput allocates dedicated model units to guarantee steady-state inference performance and strict SLA compliance for mission-critical production applications with predictable traffic volumes.

How does Bedrock model evaluation work for custom and foundation models?

Bedrock Model Evaluation supports automated algorithmic scoring across metrics like accuracy, robustness, and toxicity, as well as human-in-the-loop review teams via Amazon SageMaker Ground Truth integration.

How are Bedrock API calls secured with AWS KMS and PrivateLink?

Customer Managed Keys (CMKs) provide customer-controlled envelope encryption for model weights and Knowledge Base vector stores. AWS PrivateLink VPC endpoints keep inference traffic strictly within the customer's private virtual cloud.

How does Bedrock integrate with AWS Lambda and Amazon API Gateway?

Bedrock agents invoke AWS Lambda functions to execute enterprise actions, while Amazon API Gateway provides secure HTTP/REST endpoints with Cognito JWT authentication, usage tiers, and DDoS protection.

What telemetry and observability tools are available for Bedrock?

Amazon Bedrock streams runtime metrics (latency, token consumption, error codes) directly to Amazon CloudWatch, writes complete invocation audit records to CloudWatch Logs or S3, and integrates with AWS X-Ray for distributed agent tracing.

How can organizations implement chargeback and cost allocation for Bedrock?

Organizations apply AWS Cost Allocation Tags (e.g. CostCenter, Team, Environment) to inference profiles and application roles, analyzing usage via AWS Cost Explorer and establishing budget guardrails with AWS Budgets.

What chaos testing and disaster recovery patterns apply to Bedrock?

Resilience patterns include Route 53 Application Recovery Controller (ARC) regional failover, AWS Fault Injection Simulator (FIS) synthetic throttling injection, and circuit breakers with fallback response caches.

Can I earn an official certificate upon completing AWS Bedrock Academy?

Yes! Completing all 25 modules unlocks the official Infinity AWS Bedrock AI Architecture Mastery Certificate with your cryptographic verification ID and one-click LinkedIn badge sharing.

Escalation governance is most important when:

Risk and impact are high — High-impact workflows require clear human decision points and accountability.

Human Oversight and Escalation Governance — Compliance and Governance

← Back to Dashboard

1. Policy Mapping and Control Evidence2. Human Oversight and Escalation Governance

Human Oversight and Escalation Governance

📚 Compliance and Governance⏱ 10 min⭐ 105 XP⌨ Hands-on lab

Keep Humans in the Right Loop

Define explicit oversight triggers for sensitive decisions, legal risk, and high-impact automation. Human oversight is not a fallback of failure only; it is a design requirement for regulated workflows - and "a human reviews it" only counts if that human has time, context, and authority to actually intervene.

Choose the Loop Deliberately

Pattern	How it works	Right for
Human-in-the-loop	AI proposes, human approves before execution	Irreversible/high-impact actions: payments, terminations, legal notices
Human-on-the-loop	AI acts autonomously; humans monitor dashboards and can halt	Medium-risk, high-volume: ticket routing, draft replies
Human-out-of-the-loop	Fully autonomous within tight bounds	Low-risk, reversible: tagging, summarisation, caching

Trigger Criteria That Actually Fire

escalate_to_human if:
  action.impact == "irreversible"          # deletes, payments, external comms
  or confidence < workflow.threshold        # model uncertainty
  or guardrail.intervened                   # safety signal
  or request.matches(legal_risk_patterns)   # regulatory surface
  or tenant.tier == "regulated"             # contractual requirement

Design Against Rubber-Stamping

Give reviewers evidence, not homework - the approval screen shows the proposed action, the reasoning trace, and the top evidence links; not a raw transcript.
Meter reviewer load - an approver handling 400 requests/day is a click-through machine, and your "oversight" is fictional. Track approval latency and override rate as health metrics.
Log the human decision - approvals and overrides are audit events with named principals; override patterns feed back into routing thresholds.
Authority must be real - the reviewer can block, modify, or escalate without a VP sign-off chain that everyone bypasses under deadline.

Metric that exposes theatre: if override rate is ~0% for months, either the AI is perfect (unlikely) or oversight has decayed into rubber-stamping. Sample-audit approved items to find out which.

⌨ HANDS-ON LABConfigure Amazon Augmented AI (A2I) Human Review Loop

⭐ +150 XP

Wire low-confidence model predictions to human review workflows via Amazon A2I human review loops.

1Describe the human review workflow configuration.

lab-sandbox — simulated environment

INFINITY LAB SANDBOX v2.6 — simulated shell
Type the command for the current objective. Helpers: "hint", "solution", "clear".
$

OBJECTIVE 1 / 1 — type "hint" if stuck

🧪 Knowledge Check

Press 1-4 to select1 of 2

Human oversight should be:

Ad-hoc and undocumented

Explicitly designed with trigger criteria

Removed for speed

Only for demos