Questions bankers and allocators actually ask.

A nightly reader of SEC filings. Every night it reads what hit EDGAR's daily index for roughly 676 US-listed issuers with a public float between $15 million and $400 million, pulls each issuer's structured financials, scores who shows the signals that precede a capital raise, and writes a one-page dossier per name with every number linked to the filing it came from. Subscribers get the ranked brief by email at 5:00 AM Eastern, before the market opens.

People whose job is finding small public companies that need capital before the competition does: placement agents, equity capital markets and healthcare investment bankers, origination desks, and the PIPE, structured-equity and convertible-debt investors who write the checks. If you have ever wished you had called a CFO three weeks before the 8-K, it is for you.

• Runway: months of cash left, from XBRL cash and burn, with the age of the cash figure shown so a stale number is never mistaken for a fresh one.
• Structural need for equity: going-concern language, listing-deficiency notices (especially the stockholders' equity standard, which cannot be cured with debt), reverse splits.
• Readiness to transact: a fresh S-3, a baby shelf and its remaining capacity, an active at-the-market facility, a prior unregistered sale.
• Capital-structure pressure: convertible notes maturing inside twelve months, variable-price converts, acceleration events.
• Distress and governance: late filings, auditor changes, officer departures.
• Sector timing: catalysts such as a clinical readout that pull a raise forward.

The weights are published in full on the rubric page.

A ranked list, ten names by default, with a one-page dossier for each: the score and its components, the filing evidence with links, runway with the as-of date, shelf status and capacity, the instrument the issuer is most likely to use, and a CFO-first contact block that is verified verbatim against the filing before it is printed. The brief goes to the subscriber, and up to two colleagues at the same firm can be copied on every brief.

Filings are read the night they appear in EDGAR's daily index, so a brief reflects everything filed through the prior business day. Structured financials come from the SEC's XBRL frames and can lag the filing by a day or two; the dossier shows the date of every figure so you can judge it.

Two things, reported separately on the evidence page. Recall: of the financings announced in a twelve-month window, how many had the issuer flagged in advance, and how many days early. Precision: of the days an issuer first crossed the surfacing threshold, how often a financing followed within 90 and 180 days, against a base rate for issuers the engine never flagged. Both are computed point in time, scored only on information that existed on the day, and both are floors, because no language model ran in the replay.

Because a catch rate you cannot check is a marketing number. Every financing in the window is in the table, including the ones the engine did not see coming, and every row links to the filing. Filter it to your sector and count for yourself.

Because it only reads public filings. An issuer that raised with no going-concern language, no shelf, no maturing convert, no listing notice and no late filing gave the engine nothing to read. Some raises are genuinely invisible from EDGAR alone, and the honest answer is that no filing-based system will catch them.

Unknown until we run them. Anyone who quotes you a hit rate on your names before seeing them is selling. What we will commit to is the measurement: a replay against thirty to fifty names you have actually financed, with the lead times you can check against the filings yourself. That calibration call is included with every seat.

Those screens rank on price, volume and reported fundamentals. This ranks on the filing events that precede a raise, restarts the clock at each issuer's previous financing so chronic distress is not mistaken for a fresh signal, reports a measured lead time, and explains each name in a page you can forward. It is a complement to the terminal, not a replacement for it.

No, and no. The engine ranks issuers and explains why they surfaced. It does not recommend, size, or time any transaction, and nothing in the brief is a recommendation to buy, sell, finance, or refrain from any security or transaction. The operator is not a registered investment adviser, broker-dealer, or FINRA member, and the brief is not a research report. The terms say this at length.

Never. Public SEC filings on EDGAR and public exchange disclosures only. There is no data subscription, no expert network, no web-scraped alternative data, and no channel by which non-public information could enter the pipeline. That is a design constraint, not a policy.

No. The engine reads filings and writes dossiers. Whether and how to contact a CFO is entirely the subscriber's decision and responsibility.

$1,500 per month, one price for every seat. Whether you run deals or write checks, the price is the same; what changes is how the rubric is calibrated to your desk.

Fourteen days free. A card is collected at signup through Stripe and nothing is charged for fourteen days. Cancel before day fourteen and nothing is ever charged. After that the seat renews monthly at $1,500 until you cancel. The first brief arrives the next morning at 5:00 AM Eastern.

From the link in every brief, from the Stripe receipt, or by emailing billing@mvat.ai. Cancellation takes effect at the end of the current billing period and the brief keeps coming until then. There are no cancellation fees.

Up to two colleagues at the same firm can be copied on every brief at no extra cost. A seat covers a desk of three. Forwarding beyond that, redistributing, or reselling the brief is not permitted under the terms.

Not at the moment. One price, month to month. If you need something different, email hello@mvat.ai; the AI twin cannot negotiate and will tell you so.

At signup you answer how you usually participate (leading PIPEs, writing converts, taking down shelf paper, accumulating then negotiating, or a mix), your sectors, exchanges and float band. That selects a preset of rubric weights. On request, Ariel runs the engine against thirty of your own names and adjusts the weights from what it finds. The rubric stays published either way.

Ariel Agor, who modeled structured-finance cash flows at Deloitte and now builds production AI systems. He wrote the engine, its backtest, this site and the phone line himself. A New York family office was the first user; it is not named, by agreement.

An AI twin of Ariel, on (775) 252-8333, any hour. It says so at the start of every call. It can explain the product, answer these questions, take your details, email you the secure trial link and confirm the signup while you are on the line. It cannot quote a price other than the published one, offer a discount, promise a refund, or take a card number by voice.

It is stored, it is emailed to a person, you receive one acknowledgement, and the next morning's brief comes to you as a sample. No list, no sequence. The privacy notice covers the rest.