Hunter Brennick AI Systems & Advisory ↗
AI Systems Orchestration
Part I · Foundations/Chapter 03
03

Compound AI Systems

The shift from single model calls to systems: retrieval, code execution, verifiers, sampling, model programs, and feedback loops.

4 min read1 figures
The gist
  • AI progress often comes from systems around the model, not the model alone.
  • The best system is rarely the most agentic. It is the one with the strongest feedback loop.
  • "Ask the model harder" is one lever. Search, verification, tools, and state are the others.
ONE TICKET · THREE SYSTEMS 1 · ONE SHOT ticket modelclassify from text alone verdictbilling or not cheapest · fastest least reliable on ambiguous tickets 2 · PROMPT + TOOL CALL ticket modelneeds account state account lookupone extra call modelgrounded classify verdictaccuracy gain 3 · SAMPLE AND VOTE ticket classification 1 classification 2 … five in total majority votetake the consensus verdictsteadier five calls · use when mistakes are costly
FIG 03.1One ticket three ways: the same classification as one shot, tool grounded, and sampled with a vote.

Module 3: Compound AI Systems

This is the important shift: AI progress often comes from systems around the model.

Several real systems show what this shift looks like in practice.

Retrieval plus model is Google Search. It combines several pieces:

  • retrieval
  • ranking models
  • a knowledge graph
  • a generated summary layered on top

No single model call replicates that.

Model plus code execution shows up in a coding assistant that writes a function and then runs the test suite against it instead of asserting the code works.

Sampling plus scoring is how AlphaCode 2 operates. It generates on the order of a million candidate solutions per problem. It then uses a separate model to filter them down to a handful worth submitting.

Generator plus verifier means one model produces an answer while a second model checks it. Even basic retrieval augmented generation improves once a second pass verifies the output is actually supported by the retrieved source instead of trusting the first draft.

Planner plus executor splits the work. One model determines the steps. Another executes them.

Symbolic tools plus LLMs means delegating to a calculator or a database query or a unit test instead of asking the model to compute the answer itself. Code already does that reliably.

Model programs optimized against metrics replace hand tuning a prompt by feel. Evaluate it against a scored test set and let the results decide what works.

Build by solving one task three ways. Take a simple classification: is this support ticket a billing issue or not.

One shot classifies directly from the ticket text. It is the cheapest and fastest approach and the least reliable on ambiguous tickets.

Prompt plus tool call gives the model an account lookup tool before it classifies. That costs one extra call but produces a meaningful accuracy gain whenever account status actually matters to the decision.

Sample and vote generates five classifications and takes the majority. Five calls buy one or two extra correct answers. That is worth it only when an incorrect classification is costly.

Compare these across all three before deciding which one fits:

  • quality
  • cost
  • latency
  • failure modes

The best system is often not the most agentic one. It is the one with the strongest feedback loop. Asking the model harder is only one lever. Other levers worth pulling first include:

  • search
  • verification
  • tools
  • state

One task run three ways teaches more than a month of framework tourism.

Next chapterChapter 04 · Agent And Workflow Patterns The pattern language: workflow vs agent, prompt chaining, routing, parallelization, orchestrator workers, and evaluator optimizer loops.