← SkillSafe / PortCo Desk

Where is this company against the plan it was underwritten on?

Paste the pack, get the variance read with each material gap classified as timing or trend.

One reporting pack, four lanes. Your browser does every piece of arithmetic first — variances, unit economics, MOIC, an IRR solved by bisection, the value-creation bridge and a five-by-five exit sensitivity grid — free, with no account. The model is then held to those numbers and asked for the judgement arithmetic cannot give.

The first example ships with a saved model run for all four lanes, so you can see every lane in full without signing in and without spending a credit.

nothing pasted yet
Drag a CSV or TSV export in, or Everything is read in your browser. Nothing uploads until you run a lane.
no deal terms yet

One key: value per line. Give any two sides of the entry triangle — EV, EBITDA and multiple — and the browser completes the third; give an exit multiple and either an exit EBITDA or an EBITDA CAGR and it computes the whole return.

Paste a pack to price the run.

What this does, and what it does not

The prescan is a real table reader and a real calculator, not a keyword search. It decides the delimiter on evidence across the whole paste rather than off the first line, resolves which column is the actual, the plan and the prior year from the header — and says so when it had to assume instead. It reads figures the way packs actually write them: (1,234) is negative, $1.2m is 1,200,000, 8.5x is a multiple, 1 234,56 is a European decimal. It knows over fifty standard line items whatever they are called, and it knows which direction each one is supposed to move — so a churn line 8% under plan is favourable and a revenue line 8% under plan is not, and a percentage line is varied in percentage points rather than as a percentage of a percentage.

The returns arithmetic is arithmetic. Entry EV, multiple, equity and net debt complete each other wherever two of the sides are known and the derivation is stated. IRR is solved by bisection over the annual flow vector and the iteration count and the residual are shown, because a rate quoted without saying how it was found is a rate you cannot check. The value-creation bridge attributes the equity gain to EBITDA growth, multiple change and debt paydown, and reports the unexplained residual instead of quietly absorbing it into one of the three. The sensitivity grid computes every cell rather than interpolating between corners.

What it does not do: estimate. A ratio whose inputs are absent comes back as not derivable, named alongside the line item that would unlock it, and the model is forbidden from filling the hole with a plausible market number. It has not seen the company, only the pack. It is not a valuation, not investment advice, not a fairness opinion and not an audit, and it never tells you to buy, sell or hold anything.

Nothing to hand? Load the — revenue behind plan, EBITDA ahead of it on deferred hiring, churn at twice the underwritten rate — or the , beating its EBITDA budget while inventory, capex and scrap all run over it. The first replays a saved run in all four lanes for free.