SEGMENT LEVERS

MID-MARKET
UPPER MID
ENTERPRISE
Consumption, year 1 base £/mo
Idle cash per customer £M
FX volume per customer £M/yr
Flow share at start %
Idle share ramp pp/mo
Idle share cap %
FX share ramp pp/mo
FX share cap %

PRICING AND MARGIN

Consumption ramp capmultiple of year 1 base
Months to reach cap
Idle cash yield, bps
FX routing, bps
COGS, % of consumption

COSTS AND FUNDING

Contingency on costs, %plan carries 10% on every line
Operating costs vs plan, %scales every month's opex
Seed, $MSep 2027
Series A, $MMar 2029
Freeze costs afterall costs and hiring held at that year's run rate, sales keeps selling at that pace
Customer adds, churn, headcount and the opex schedule come from the model and are held fixed, except the opex scale. Everything else recomputes live. Free trial is three months per customer. Flow share ramps on the plan's calendar basis.

Intelligence is the next
financial infrastructure.

Signl is the governed layer between financial data and every agent, application and financial decision.

This is a live plan, not a deck. Every number comes from the financial model. On the last three pages, move a lever and everything recalculates.
Tom Greenwoodtom@signl.ai
Matt Komorowskimatt@signl.ai

YEAR 1 · THREE FREE MONTHS, THEN RAMPING

Reasoning is consumption at the base price, ramping with usage. Yield is 20bps on the idle cash Signl routes, FX is 5bps on the volume it routes. Contribution is after COGS and deployment cost.

Treasurers told us the same thing, in their own words.

Discovery interviews with Group Treasurers and CFOs, August to September 2026.

Verbatim from recorded discovery calls, lightly edited for length. PepsiCo, Booking.com, Flutter Entertainment, Imperial Brands, Remitly, Deel, Flywire, Bolt, Evervault, Money Dashboard and others.

Where it goes from here.

Win the beachhead and own the decision. Own the decision and everything downstream is yours to optimise.

NOW
Liquidity on the governed model
One shared model of cash, FX, yield and risk. Every number carries object, source, confidence, dependency and consequence. Decisions and actions run through policy and audit.
Cash positionForecastFX routingYield routingRiskApprovals
PRODUCT
Extends what already exists
Same model, same agents, same controls. New surfaces and new users.

Beyond the CFO

Sales, supply chain and operations reason from the same governed model, with the same evidence, confidence and controls.

Unlocks: business intelligence with financial meaning built in

Settlement and payouts

PSP and acquirer settlements explained and matched to bank and ERP, with fees, reserves and timing on one model.

Unlocks: the money in flight for every company that takes payments

Treasury management system

Cash, debt, hedging and payments for mid market companies with no treasury system today, on the same model.

Unlocks: a second product on the same model, sold to a new buyer

Substrate as infrastructure

Third party agents reason from the governed model through MCP, with Signl policy and audit on every call.

Unlocks: platform revenue, with every lab’s agent as distribution
REGULATED
Needs licences and capital
Signl stops routing and becomes principal. The spread, not the referral fee.

Principal FX

Signl becomes the counterparty rather than the router, pricing and settling the FX it already sees and recommends.

Unlocks: 5bps on routing becomes 20 to 50bps on the same flow

Principal yield

Signl holds the money market fund rather than routing to one, so the idle cash it finds earns on Signl’s own balance sheet.

Unlocks: 20bps on routing becomes a full spread on AUM

Payments

The execution adapter becomes the rail. Approvals and audit already exist, so payments run on the governed path.

Unlocks: payment volume and fees on a governed, approved path

Stablecoin settlement

Signl settles on chain with the approvals and audit already in place, cutting the cost and time of moving money across borders.

Unlocks: cheaper, faster settlement for payments and FX
Sequenced top to bottom, after liquidity is proven. Each row uses everything the row above has already built.

Liquidity is the wedge. The financial state is the platform.

Each new workflow reuses the same evidence, relationships and authority.

NOW / LIQUIDITY
What cash is actually available?

Position, restrictions and obligations, forecast and headroom.

SIGNL SUBSTRATE
One governed financial state

Cash and flows, entities and obligations, evidence and confidence, policy and authority.

Every number carries its source, its definition and who can act on it.

TREASURY
How should we fund and deploy cash?

Funding and concentration, FX and yield decisions, policy and risk.

CFO
What does the financial state mean for the business?

Working capital, plan vs actual, scenarios and capital allocation.

Expansion direction: more decisions over one model, with evidence and control carried forward.
An enterprise system of connected sources, governed meaning, shared workspace, controls and audit. The model is one box.

    SOURCE SYSTEMS Banks ERP TMS Billing HRIS and payroll Market and FX data Payments SIGNL SYSTEM BOUNDARY Workspace Shared position, work queue, approvals and reporting for the finance team.Reads the same governed model as every agent and every application. The CFO and the agent never see different numbers. Integration Connectors Normalisation Change detection Source mapping Kept live as sources change Signl agents the built in finance team Monitor Investigate Forecast Recommend Orchestration and tool calls queries governed meaning only, never raw data Evidence and confidence checks every output carries source and confidence Trust agent Signl agent FX agent Cash agent Risk agent Execution agent Model gateway Anthropic OpenAI Open weights External. Interchangeable. Routed by task, cost, residency. What a lab sells Policy and action Permissions Approval workflow Execution adapters Outcome tracking Nothing executes unapproved recommendations Substrate governed financial meaning layer. Every number carries object, source, confidence, dependency and consequence. Object Source Confidence Dependency Consequence What the thing is registry and definitions Who owns the truth authority and lineage How sure we are freshness and agreement What depends on it graph of relationships What it changes downstream impact facts in governed context outcomes update meaning Audit ledger Append only. Every source, calculation, model call, approval, action and external query is recorded. Spans every layer above. UI MCP API PEOPLE CFO Finance Treasury Audit AGENTS Customer agents Third party agents APPLICATIONS Forecasting Reporting Modelling Claude with an MCP into the ERPOne system. Raw data. No agreed definitions.No connectors kept live. No approvals. No audit trail.No shared workspace. Nothing can execute.It can read a balance. Nobody has agreed what the balance means. One agentone playbook, one questionreads raw data, nobody will let it actOne agent built for one workflowWorks in the demo. Reads one source, runs one playbook, answers one question.Breaks the day the ERP changes a field. Cannot approve. Cannot execute. Cannot be audited.The second workflow needs a second agent, a second integration and a second argument with finance.Finance will not let it near money without everything else on this page. WHAT ACCUMULATES IN SUBSTRATE The software can be copied. The accumulated context cannot. Every connection, document, decision and correction teaches Signl how the business behaves financially. TRUST MEMORYRELATIONSHIP MEMORYJUDGEMENT MEMORYOPERATING MEMORY How this business treats confirmed,probable and uncertain information How entities, obligations, exposuresand risks interact across the group Which approvals, overrides and evidenceshaped each decision taken How work is routed, approved,escalated and resolved across the team A replacement cannot import balances and declare equivalence. It has to reconstruct every definition, dependency, permission and precedent and prove they survived. A competitor arriving later does not inherit that authority. It starts at observation. This is switching cost, and we say so plainly. It does not win customers, but it does help us keep them.

    LANDSCAPE

    TMS and ERP hold the records. BI shows them. The labs sell the model. Nobody governs what a number means before an agent acts on it. That gap is the layer Signl occupies, between the systems and everyone who reasons on them.

    WRAPPER

    Yes, it is getting easier to point a model at one system or build one agent for one workflow. Everything else on this page is what finance needs before software goes near money.

    CATEGORY

    Financial intelligence. The governed path from financial data to financial decisions. Everything inside the boundary is the category: connect, govern meaning, reason, approve, act, audit. Not a dashboard, not a chatbot, not a TMS, not a model.

    MOAT

    Substrate. A governed model of each company’s financial meaning, built with the customer. It compounds with every decision, correction or action. The customer owns its data. Signl owns the context model built from it. The data is portable. The context model is not.

    Hover any block in the diagram for what it does and why it matters. Hover a card below to see which parts of the system answer it.

    COHORT ECONOMICS · EACH YEAR'S NEW CUSTOMERS, FOLLOWED THROUGH 2031

    REVENUE BY COHORT · £M PER MONTH

    CASH CURVE BY COHORT · £M

    WHAT THE COHORTS EARN IF COSTS ARE FROZEN