Two Databases. One Twin.

The Three-Engine Core.

Network infrastructure has two distinct states that are fundamentally different in shape: intent (what should be true) and reality (what is true right now). Mixing them in one schema creates ambiguity that compounds over time. A Go control plane routes every query across three specialised engines in one execution pipeline, so neither state is forced into the wrong shape and multi-table JOIN explosions never happen.

The relational state engine (Postgres) is the deterministic ledger: CIDR allocations, static reservations, VLAN metadata, ownership and policy, under ACID guarantees.

The semantic search engine (pgvector with an embedded local model) embeds incoming syslogs, BGP events, and device configs in milliseconds inside the control plane. Natural-language search across historical logs runs with no third-party API dependency and no external token fees.

The spatial graph engine holds reality: live topology, physical connections, observed state. Index-free adjacency computes multi-hop relationships (chassis ports to switches to VLANs to DHCP scopes to DNS records) in microseconds, which is what makes a live interactive canvas and sub-millisecond blast-radius analysis possible.

The agent runtime sits across all three, continuously comparing intent to reality. When they diverge, that's drift, and drift is the fundamental signal the platform acts on.

Streaming, Not Request/Response.

The platform is gRPC/MCP-streaming-first rather than REST. When an agent executes, the UI receives a stream of typed events - status updates, data chunks, tool calls, and a final result - rather than waiting for a single response payload.

This means the engineer sees the agent's reasoning as it happens: which vector matches were found, which tool the agent chose to invoke, and whether the final action was committed or deferred for approval. No spinner. No black box.

// Stream of typed events, not a single response
[
  status       { "type": "agent_status",    "body": "searching..." }
  vector match { "type": "data_chunk",      "body": { "matches": 3 } }
  tool call    { "type": "agent_tool_call", "body": { "fn": "allocate" } }
  commit       { "type": "final_result",    "body": { "committed": true } }
]