Lesson 7 — Integration: The Bridge of Systems
Learning Outcome: Connect multiple brains into a cooperative mesh; learn how orchestration, modular APIs, and distributed reasoning create collective intelligence.
Manual
The Bridge Between Minds
A single brain can reason; a network of brains can coordinate. Integration is not about building a bigger model — it's about weaving interfaces that let many smaller minds exchange context.
1 · Modularity and Message
Each module specializes: retrieval, summarization, generation, evaluation. They communicate via messages — structured payloads describing intent, content, and confidence. Meaning moves as protocol, not as authority.
2 · The Orchestration Layer
The orchestrator routes messages between brains. It decides which module should respond next and merges outputs into a unified answer. This is synthetic dialogue among intelligences.
3 · Chaining and Relaying
The Foundry supports chains and relays:
/chainsequences calls (A → B → C)/relayallows dynamic decision-making (A asks B or C based on confidence)
Each link defines transformation, not hierarchy.
4 · Emergent Understanding
When feedback flows both ways, local reasoning becomes collective awareness. The system starts noticing its own processes — an echo of meta-learning.
Workshop
Hands-On: Chain Two Brains
Goal: build a minimal multi-brain relay using the Foundry endpoints.
Multi-agent orchestration demo placeholder
(Full interactive demo available at brainfoundry.ai)
Manual Steps
1️⃣ Define two brains
- Retriever: performs
/search - Summarizer: performs
/respond
2️⃣ Register the chain
curl -X POST /api/foundry/chain \
-H "Content-Type: application/json" \
-d '{
"chain_id": "lesson7-demo",
"steps": [
{ "role": "retriever", "endpoint": "/search" },
{ "role": "summarizer", "endpoint": "/respond" }
]
}'3️⃣ Send a relay query
curl -X POST /api/foundry/relay \
-H "Content-Type: application/json" \
-d '{ "chain_id": "lesson7-demo", "input": "Summarize the knowledge about orchestration across systems." }'Observe how the orchestrator automatically calls the retriever first, then the summarizer, merging results.
Reflection
Reflection — Shared Cognition
How does understanding change when it becomes shared? Write 5–8 sentences imagining two minds (human or synthetic) cooperating on a problem. What do they gain — and what do they lose — by thinking together?