V1 Engine

An engine, not an agency

What it is

An innovation agent, not another copilot

Innovation shouldn't be a privilege of headcount, budget, or permission. The V1 Engine is an AI agent purpose-built for innovation — deployed in your business, connected to your systems, governed by your rules. It turns anyone's idea into a working, approved, shipped thing. That's what innovation looks like when it's ordinary work.

General assistants answer questions. The V1 Engine ships innovations. The intelligence is the commodity. The method, the connections, and the governance are the product.

How it works

Five layers that turn ideas into products

The architecture separates what is commodity (model intelligence), what is V1's IP (the method, orchestration and gateflow), and what belongs to you (data, systems, infrastructure and everything shipped).

Surfaces

FOR EVERYONE IN THE BUSINESS

Where people meet the Engine. A web workspace plus the places work already happens.

Web appSlack / TeamsEmbedded in internal tools

Agent core

V1 IP

A frontier model wrapped in the V1 method layer: encoding how to find, shape, evidence, and ship product. The model is swappable commodity; the method layer is the product.

Frontier model (swappable)V1 method layerImpact assessorSession + decision memory
Tool calls · read-only · least privilege

Connector layer

CONFIGURED PER CLIENT

Typed connectors that let the agent reason about your real context. All read-only. Scoped to your tenancy.

DATA: analytics · insights · warehouseSTRATEGY: strategy · personasARCHITECTURE: cloud · deployment · logsUX: UI components · standardsBRAND: assets · copy systems

Your systems

CLIENT-OWNED

The stack you already run. The Engine reasons against what exists — no new tracking, no copies.

Warehouse / BIAnalyticsCloud + CI/CDDesign systemDAM / brandDocs + knowledge
Writes happen only through the gate

Deployment runtime

YOUR INFRA · V1 OPERATED

The only path to production. Builds deploy into your environment after clearing the gateflow. Full audit trail.

Gateflow engineSandbox / pilot environmentsProduction deployAudit + telemetry
■ V1 IP ■ CLIENT-OWNED ■ COMMODITY
What it's made of

Six components, one system

AGENT CORE

The method, running

Orchestrates every build. Knows what matters because it's tuned on shipped work — which evidence to trust, which questions to ask, when an idea is an optimisation versus a bet.

CONNECTOR LAYER

Context, not copies

Read-only connections into your stack. Proposals arrive on-strategy, evidenced, buildable, and on-brand. No generic outputs.

ARTIFACT ENGINE

Never a blank canvas

Generates everything an idea needs: prototypes built to your UI, business cases against your data, pricing implications before anyone commits.

GATEFLOW ENGINE

The control plane

Assesses impact, assigns a tier, and routes it — straight deploy, single approver, or full sign-off. Configured to your existing governance.

DEPLOYMENT RUNTIME

Ship inside the perimeter

Sandboxes for pilots, production promotion in your cloud, full audit trail. Nothing reaches users except through this path.

TELEMETRY

Accountable to results

Shipped things report back: usage, conversion, retention. The Engine learns from what actually happened, not what was predicted.

From idea to shipped

Every build moves through six stages

01

Describe

Anyone states a problem or idea in plain language, from any surface.

02

Build

The agent and the person iterate to a working artifact: prototype plus evidence pack.

03

Pilot

Runs live with a controlled group in a sandbox. No approval needed to get here.

04

Gate

Impact assessment assigns a tier and routes to your configured approval path.

05

Ship

Deployment runtime promotes to production inside your environment.

06

Measure

Telemetry feeds results back into decision memory. The Engine gets smarter.

Tier
Trigger
Examples
Route
Low
No revenue, customer, security or data implications.
Internal tools, FAQs, process aids, content drafts.
Straight deploy
Medium
Customer-facing or brand-touching, no money or sensitive data.
Experience changes, comms, marketing experiments, pilots.
Single approver
High
Revenue, pricing, security or data implications.
New propositions, pricing changes, anything moving money or PII.
Full sign-off chain
Designed for security

Built to pass your hardest review

  • In-environment deployment: your cloud, your tenancy, your perimeter
  • Connectors are read-only, least privilege, individually scoped and revocable
  • Writes occur only through the deployment runtime, only after the gate
  • No new tracking in your product; reasoning on patterns and aggregates
  • Full audit log on every build, gate decision, and deployment
  • Client data never trains shared models
Configured per deployment

Your business, your rules

  • Gate tiers, thresholds and approval chains mapped to your governance
  • Connector scope per system: what the Engine can see and use
  • Your UI component library and web standards
  • Your brand assets and copy systems
  • Deployment targets and sandbox policy
  • Surfaces: which channels the Engine is reachable from

Everything else — the agent core, the method layer, the gateflow engine — is the same product everywhere. Every deployment benefits from every other one.

Two ways to use it

A subscription, or a service provider.
Either way, you own the IP.

Innovation as a Service

A subscription. Your teams get the Engine deployed on your infrastructure with the tools, methodology, and playbooks to innovate safely. You own the IP.

Innovation as a Service Provider

Fixed-cost retainer. We ship, you own the IP. Full-service delivery — you brief the opportunity, our crew builds the innovation. The returns compound.

Your enterprise
needs this engine