AI Incident Timeline: Build a Reliable Summarizer
This senior-level guide covers AI incident timeline with production trade-offs for reliability, security, and cost.
prompt change management: Change Management for Prompts
This senior-level guide covers prompt change management with production trade-offs for reliability, security, and cost.
AI agent runtime: Build vs Buy the Agent Runtime
This senior-level guide covers AI agent runtime with production trade-offs for reliability, security, and cost.
Cursor Cloud Agents (formerly Background Agents): CI Gates That Catch Hallucinated AWS Changes
CI gates for Cursor Cloud Agent pull requests on AWS CDK: blast-radius limits, tsc and cdk synth, cfn-lint, a template gate for wildcard IAM and AdministratorAccess, CODEOWNERS and Cursor project rules.
Posted inAI AWS Machine Learning
Bedrock Agents + EventBridge: Async Tools Without Blocking the Session
Bedrock Agents + EventBridge: offload long tools with start/status APIs, correlation IDs, and fail-closed IAM so sessions stay interactive while workers run the slow path.
Posted inAWS System Design
EventBridge Pipes to Lambda: Enrich, Filter, and Fail Closed
EventBridge Pipes to Lambda done safely: fail-closed filters, idempotent enrichment, partial-batch failures, and DLQs — so agent and order pipelines do not silently drop events.
CloudWatch Logs Anomaly Detection: Ignore Deploy Noise, Catch Novel Errors
This senior-level guide covers CloudWatch Logs Anomaly Detection with production trade-offs for reliability, security, and cost.
Posted inAWS System Design
API Gateway WebSocket Fan-Out: Coalesce Pushes After Reconnect Storms
This senior-level guide covers API Gateway WebSocket Fan-Out with production trade-offs for reliability, security, and cost.
X-Ray Service Maps Lie: Fix Missing Spans in Node Trace Edges
This senior-level guide covers X-Ray Service Maps Lie with production trade-offs for reliability, security, and cost.
Posted inAWS Machine Learning System Design
Feature Stores vs Inline Compute: Real-Time Decisions Without Training Skew
This senior-level guide covers Feature Stores vs Inline Compute with production trade-offs for reliability, security, and cost.