Forward-deployed / Learning zone
AI engineeringfor AI-native PMs
A linked curriculum

AI engineering,
from scratch to production.

The engineering discipline underneath production LLM systems — inference, retrieval, evaluation, observability, safety, and cost — taught through the lens of the decisions a Senior or Principal PM has to make.

7 modules23 lessons PM-nativeproduction-grade
🎯
Built for the AI-native PM. Every lesson pairs the real mechanics with a briefing: why it matters to the product, what it changes in your decisions, the question to ask your eng team, and the product risk if you ignore it.

The modules

00

Foundations

The mindset shift from “writing prompts” to “engineering systems.”

  • harness engineering
  • context engineering
  • infra not demos
◆ interactive demo + recap Open module →
01

Inference internals

What happens between your request and the tokens that come back.

  • prompt vs semantic caching
  • kv cache management
  • prefill vs decode
  • +3 more…
◆ interactive demo + recap Open module →
02

Reliable outputs & tool use

Making models produce things downstream systems can trust.

  • structured output
  • function calling
  • agent guardrails
  • +1 more…
◆ interactive demo + recap Open module →
03

RAG & retrieval

Grounding models in your data — and proving they actually used it.

  • rag architecture
  • retrieval evals
◆ interactive demo + recap Open module →
04

Evals & observability

You cannot operate what you cannot measure.

  • evals
  • observability
  • cost attribution
◆ interactive demo + recap Open module →
05

Safety & multi-tenancy

Keeping tenants, users, and data from leaking into each other.

  • safety engineering
  • multi tenant isolation
◆ interactive demo + recap Open module →
06

Strategy & tradeoffs

Picking the right tool, and naming the cost of every choice.

  • finetune vs icl vs rag
  • inference stack tradeoffs
  • production failure modes
◆ interactive demo + recap Open module →

Threads — follow one concern across the whole stack

💰 Caching & cost

How spend and latency hide in the inference layer — and how a cache can become a privacy incident.

🛡️ Reliability

From valid JSON to bounded agents: making stochastic output safe for downstream systems.

⚡ Latency

Why “it feels slow” has two different causes, and which lever fixes which.

🎯 Quality

Context → retrieval → grounding → evals: where trustworthy answers actually come from.

🧰 Which tool?

Fine-tune vs. ICL vs. RAG vs. distillation — the highest-leverage early decision.