llmfnlayers

From Prompt to Product

The interesting part of an LLM product is not the model. The model is the middle. What surrounds it — how you retrieve, structure, ground, remember, act, and measure — is what separates a demo from something that works in production. This site is a map of those eight layers.

The Progression

Reading Guide

Workshop participants

Coming to the workshop? Read layers 1–4 before the day. Each page takes about ten minutes.

Exploring on your own

Exploring on your own? Start anywhere. Each page stands alone and links to the next.

Five Stages

0 LLM as Chat Judgement
1 LLM for Software Reviewing
2 LLM Augmented Software Integration
3 LLM Integrated Application Architecture
4 Domain Specialization Expertise
Read the full article →

Implications

Ability What the model can and cannot do. Not in your control. Know the limits before you build.
Behaviour How you shape it — through prompts, context, schema, and evals.
Code The application layers — retrieval, state, memory, tools. Most teams start here. Most should start at B.
Design How humans experience and correct the system. When your LLM is wrong, does the user know?
Economics Whether it is worth running at scale. Token cost, latency, and the unit economics of one successful interaction.