The spatial memory your robots query in plain English.

Where we sit in your stack.

ABOVE perception (we consume images, depth, and poses), BELOW high-level task planning (we serve queries to your planner's LLM). We are the scene-memory + scene-QA layer — nothing more, nothing less. That focus is the pitch.

The wedge in one sentence: a per-facility splat that lives forever + an MCP API any LLM can query — sold per-capture and per-query, hosted, no infra burden on the customer. The moat isn't the splat — it's that the LLM is not vendor-locked, which matters to robot CTOs who already have an LLM relationship and don't want to be told which one to use.

What a robot actually asks — and gets back.

The read-only scene-query API is the Phase-1 robotics focus. It answers object-relationship and path-clearance questions over the stored per-facility scene — relative position between two named objects, free space along a planned segment, who's near a waypoint — and returns structured JSON your planner acts on directly, not prose. Every response below is the real shape these endpoints return today.

The example scene is one captured world, warehouse-aisle-7 — 42 objects: pallets, racks, fixtures (a dock and two charging bays), and two people. Distances are in the engine's world_units, calibrated to the facility's metric scale at capture time. An unknown world returns 404 (existence is never disclosed); an empty world returns honest zeros, never a fabricated obstacle.

These endpoints layer over the 5 read-only scene-memory MCP tools the planner's LLM calls; the runtime returns structured JSON and your LLM and planner interpret it however they choose. Every call is metered — per-query counts persist to durable usage records, with per-pilot rollups and SLA-grade p50/p95/p99 latency + error-rate read back from /scene-query/usage/{summary,by-caller,sla}. That measured volume is what the 90-day pilot converts on: the meter, not the price — pricing stays your decision. Scene-delta ("what changed since shift T") is on the Phase-2 roadmap, not yet shipped — we don't show it as live. The MCP surface also exposes 12 audited mutation tools (anchors, sessions, transforms) — covered in the safety section below.

Safety and authority: how the runtime keeps control.

The runtime is the authority over physics, collision volumes, transforms, and the canonical scene state. An LLM assistant proposes actions through the MCP mutation tools; it does not silently rewrite the scene. Every mutation is audited.

A robot queries its facility, in one diagram.

Your robot's planner (or its LLM) talks to RakuAI's MCP surface over HTTPS — or via a stdio bridge from an on-robot Jetson Orin client. The runtime answers in spatial primitives: anchors, poses, scene queries, splat handles, object deltas.

Both halves are real and shipping. The production MCP server exposes 17 native tools (5 read-only + 12 mutation) plus a passthrough surface for LLM-specific extensions — see spatial-engine.html for the full inventory and mcp.html for the documented scene tools. The hosted MCP runs on Azure Container Apps in East US. The LLM is yours to pick — that vendor-neutrality is the moat.

Who this is for.

Four segments, one shared pain: robots collect terabytes of spatial data that get dumped into storage and never queried again. Lead with warehouse for revenue stability; humanoid for visibility.

The partner-resource stack behind a pilot.

RakuAI's NVIDIA Inception membership (accepted May 2026) and pending startup-credit programs compound here. Each is honestly labelled by status — we don't claim approvals we don't have.

The ROI math.

We're not replacing your perception stack. We're replacing the persistent-memory subsystem nobody wants to own and the scene-QA LLM bridge that gets pushed to "next quarter" forever.

Net savings, year 1: $250K–$1M in engineering cost displaced. Honest caveat: if a prospect already has a happy persistent-memory + LLM-bridge stack, we don't have a sale — and we won't pretend otherwise.

The 90-day free pilot.

Free for the full 90 days. The output is a published, co-marketed case study — the deliverable that makes pilots #2, #3, and #4 easier to land.

Conversion offer: per-facility license OR per-capture + per-query metering — we let the first 3 pilots tell us which pricing model wins. The pilot stays free regardless.

Where we sit vs. the rest of the field.

The honest version. Foundation-model robotics firms are integration partners and customers, not competitors.

If you want to go deeper before you write.

The pages robotics teams read most before a first call.