A recommendation layer for corporate treasury

Your forecasts already know how much cash is safe to move.

We read the balances, forecasts, and actuals you already have — and tell you how much is safely movable, where it should go, for how long, and what it’s worth. Every day, with an auditable confidence score. No new TMS. No execution. No integration project.

Built for treasury teams on any stack — TMS, cash-visibility tool, or structured spreadsheets.
MOVABLE — ABOVE THE LINE €11,300,000 CALIBRATED BUFFER — SIZED FROM YOUR OWN FORECAST ERRORS CONFIDENCE 91% · EXPLAINABLE
€1.1M / yr
what €50M of idle USD-equivalent cash forgoes at mid-2026 short rates*
91%
confidence scores that are calibrated from your own forecast errors — and explainable line by line
0
systems replaced, integrations required, or payments executed — you approve everything in your own stack
The problem

The question nobody can answer daily

Every treasury knows some cash is idle. What no TMS answers is the only question that matters: how much of it is safe to move, for how long, at what confidence? So minimum balances get set once, padded for comfort, and never revisited — while short rates pay 2–4%. The cost is invisible, because nobody computes it.

How it works

Three steps. None of them touch your systems.

1

Read what you already have

Balances, forecasts, historical actuals, scheduled payments — CSV exports from your TMS or ERP. No integration, no system access, no new data entry.

2

Calibrate a buffer from your own errors

The engine measures your forecast accuracy by account, category, and horizon — and sizes the liquidity buffer statistically, not by gut feel. When your business shifts, the buffer shifts with it.

3

Recommend — never execute

A daily plan of transfers with amount, duration, destination, expected benefit, and an explainable confidence score. You approve in your own systems. A deterministic engine — no LLM decides anything.

What you get

Every recommendation carries its own audit trail

  • A recommendation layer, not another TMS. Nothing gets replaced; your approval workflows stay exactly where they are.
  • Confidence you can defend. Every score decomposes into forecast accuracy, payment volatility, buffer coverage, and data quality — presentable to your CFO or auditor.
  • Near-misses included. When a move is rejected, you see why, what it would have earned, and exactly what would unlock it — an intercompany agreement, a pool target, an FX feed.
Sample recommendationCONFIDENCE 91%
Operating Account DE01  →  EUR Liquidity Account HQ
Move amount€4,200,000
Holding period12 days
Buffer retained after move€3,800,000
Incremental yield spread2.2%
Estimated annualized benefit€92,000
Why: historical balances and forecasted outflows indicate excess liquidity above the required threshold for the next 12 days. Full rationale and data lineage attached.
Near-missFR02 → HQ blocked: no intercompany agreement on file. Signing one unlocks an estimated €38,000 / yr.
Where every engagement starts

The Missed Yield Audit — free, backward-looking, on your own history

Before you trust anyone's forecast of the future, look at the past. We run the engine backwards over your last 12 months and show you what was provably safe to move — and what it would have earned.

  • You export a handful of CSV files — roughly two hours of your team’s time, no system access
  • You get back: average idle cash above a properly calibrated buffer, the yield it forgot to earn, and the largest moves you could have safely made
  • Worst case, you confirm your buffers are perfectly sized — also worth knowing
Request an audit conversation
Sample audit summary — illustrative
€742,000
yield forgone in FY25 on cash that was provably safe to move
Idle above calibrated buffer, avg€23.4M
Largest single safe move€4.2M · 12d · 91%
Safe recommendations, FY25 backtest147
Buffer over-sizing vs. calibrated need2.1×
Data & trust

Built the way treasury software should be

Deterministic, explainable, auditable

The recommendation engine is deterministic treasury logic — buffers, feasibility, move economics. No LLM sits in the decision path. Every output can be traced back to its inputs.

Your data, minimally and briefly

We work from CSV exports you choose to share — never system access or payment credentials. Data is processed in the EU and retained only as long as the engagement requires.

Recommendations, not execution

The engine proposes; your team disposes. Transfers are approved and executed in your own systems, under your own controls, or not at all.

*Illustration based on mid-2026 policy rates (Fed funds 3.50–3.75%, ECB deposit facility 2.25%), blended short rates on USD-equivalent balances. Sample figures on this page are synthetic and illustrative.

Talk to us

We’re working with a small number of treasury teams as design partners. If the question “how much of our cash is actually safe to move?” hits home — write us.