https://benjaminlee.dev/blog Solutions Architect at AWS. Writing about AI/ML, cloud architecture, cybersecurity, and quantitative research. en-us lilphd99@gmail.com (Benjamin Lee) lilphd99@gmail.com (Benjamin Lee) Sat, 20 Jun 2026 00:00:00 GMT https://benjaminlee.dev/blog/detecting-offensive-meaning-in-marketing-content-on-aws https://benjaminlee.dev/blog/detecting-offensive-meaning-in-marketing-content-on-aws Brand-safety failures rarely come from an obvious slur. They come from clean words that combine into something offensive, often through a pun in another language. This post presents a batch architecture on AWS, built on a multimodal foundation model core, that detects illegal, sexual, and offensive meaning in marketing content across English and Korean before it ships. Sat, 20 Jun 2026 00:00:00 GMT lilphd99@gmail.com (Benjamin Lee) awscloud-architectureaicontent-moderationbedrock https://benjaminlee.dev/blog/eight-rag-architectures-every-engineer-should-know https://benjaminlee.dev/blog/eight-rag-architectures-every-engineer-should-know How to choose among eight RAG architecture patterns, from Simple RAG to Agentic RAG, by matching each pattern to the specific failure mode it solves. Sun, 07 Jun 2026 00:00:00 GMT lilphd99@gmail.com (Benjamin Lee) ragllmai-agentsgenaiaws https://benjaminlee.dev/blog/ab-testing-llm-prompts-production https://benjaminlee.dev/blog/ab-testing-llm-prompts-production Prompt changes are code changes — they should be tested the same way. This post covers the mechanics of A/B testing prompts against live traffic, the metrics worth tracking, and the tooling that makes it tractable. Wed, 14 Jan 2026 00:00:00 GMT lilphd99@gmail.com (Benjamin Lee) llmai-agentsmlopsdata-science https://benjaminlee.dev/blog/agentic-ai-patterns-aws https://benjaminlee.dev/blog/agentic-ai-patterns-aws AWS has published prescriptive guidance on agentic AI patterns — from orchestrator-worker hierarchies to event-driven choreography. This post maps those patterns to real infrastructure decisions on Lambda, ECS, and Bedrock. Mon, 01 Dec 2025 00:00:00 GMT lilphd99@gmail.com (Benjamin Lee) ai-agentsawscloud-architecturellm https://benjaminlee.dev/blog/langgraph-stateful-agents-production https://benjaminlee.dev/blog/langgraph-stateful-agents-production LangGraph models agents as explicit graphs with typed state, built-in checkpointing, and human-in-the-loop support. Here is what that means in practice for building agents that actually run reliably in production. Wed, 12 Nov 2025 00:00:00 GMT lilphd99@gmail.com (Benjamin Lee) ai-agentslanggraphlangchainpythonllm https://benjaminlee.dev/blog/langgraph-agentic-workflows https://benjaminlee.dev/blog/langgraph-agentic-workflows LangGraph gives you explicit control over agent state, branching, and cycles — the things that make the difference between a demo and a production AI system. Here is how to think about it and where to start. Mon, 03 Nov 2025 00:00:00 GMT lilphd99@gmail.com (Benjamin Lee) ai-agentslanggraphlangchainpythonllm https://benjaminlee.dev/blog/aws-multi-account-architecture https://benjaminlee.dev/blog/aws-multi-account-architecture AWS recommends a multi-account strategy from the start — not just for enterprises. This post breaks down the OU structure, SCP guardrails, and Control Tower patterns that make it work without drowning a small team. Wed, 08 Oct 2025 00:00:00 GMT lilphd99@gmail.com (Benjamin Lee) awscloud-architecturesecuritydevops https://benjaminlee.dev/blog/real-time-streaming-kinesis https://benjaminlee.dev/blog/real-time-streaming-kinesis Kinesis Data Streams is the backbone of real-time data pipelines on AWS. This post covers the layered architecture AWS recommends, the tradeoffs between provisioned and on-demand modes, and the operational details that matter in production. Mon, 22 Sep 2025 00:00:00 GMT lilphd99@gmail.com (Benjamin Lee) awsdata-engineeringcloud-architecturepython https://benjaminlee.dev/blog/building-real-time-telemetry-pipelines-aws https://benjaminlee.dev/blog/building-real-time-telemetry-pipelines-aws A practical walkthrough of designing a high-throughput telemetry pipeline on AWS — from ingestion through Kinesis to enrichment in Lambda and storage in S3 and DynamoDB — with lessons learned from production deployments. Mon, 15 Sep 2025 00:00:00 GMT lilphd99@gmail.com (Benjamin Lee) awscloud-architecturedata-engineeringpython