1 min read
Notes from the gateway
A short welcome, and what this blog will cover: agents, MCP, RAG and LLMOps.
Welcome. This is where I will write about building and running LLM systems in production.
What to expect
Posts will stay close to the work I do day to day:
- Agents. Agent runtimes, memory, multi-agent orchestration with LangChain and LangGraph, and human-in-the-loop workflows.
- MCP. Turning existing APIs into agent-ready tools, and routing model and tool traffic through a single gateway.
- RAG. Ingestion, hybrid retrieval, re-ranking and checking that answers stay grounded in their sources.
- LLMOps. Inference serving, evaluation, observability and cost governance.
Reading the posts as data
The site runs on a small JSON API, so every post is also available as structured data:
const res = await fetch(`${API_URL}/api/posts?tag=mcp`);
const posts: PostMeta[] = await res.json();
for (const post of posts) {
console.log(post.publishedAt, post.title);
}
More soon.