Mimir
01Opportunities

Screen 22 million properties on what the record actually says.

The deal you want is already sitting in the record: the lots quietly reaching the same hands, the owner holding free and clear, the loan walking toward its wall. Screen 22 million properties and get there weeks before it lists.

02Screen

The research day, compressed.

The work this replaces is a person with twelve browser tabs cross-referencing a county recorder against a tax roll against a corporate registry, one property at a time. The same question asked of the resolved corpus returns in seconds, across three states, with the citation already attached.

  • Filter on recorded conditions across 22M properties
  • Results resolved to the person, so entities do not fragment your list
  • Export with citations intact
Screener — sourced signals1,284 resultssynthetic
Adjacent parcels, same hands3 lots on one block, 2 entities, one shared controllernyc_acris
Free and clearno open mortgage on record since the 2011 satisfactionnyc_acris
Long holdheld 14 years — above the block median of 6nyc_acris
Tax warrant on the ownerstate warrant filed against a controlled entityny_dtf
Open hazardous violation2 class-C violations, no release observednyc_hpd
Every signal is a recorded filing with its source attached. None of them tells you what to do.
03Assemblage

The block that is quietly consolidating.

Adjacent parcels moving into the same hands through different entities is the highest-signal pattern in the public record. It is also completely invisible parcel by parcel, which is why almost nobody trades on it.

Parcel view — controlled by one personsynthetic
HARROW STREET CAPITALKESTREL 1220BELLHOUSE EQUITIES
04Concentration

Where one owner is over-committed.

Concentration is an opportunity signal and a risk signal depending on which side of the table you sit. Either way it requires resolving the person first, because the exposure is spread across entities that share no name.

Exposure by entity — one borrowersynthetic
Same person, five entities
  • HARROW STREET CAPITAL$15.6M
  • KESTREL 1220$11.9M
  • BELLHOUSE EQUITIES$7.0M
  • VANDERPOOL HOLDINGS$4.1M
  • PELL ROW PARTNERS$2.6M

Underwritten one entity at a time, this reads as five unrelated borrowers. Match the person, and it is one.

05Timing

A saved screen becomes a standing alert.

The screen you ran today is the alert you want tomorrow. Save it and Mimir tells you when a new property meets the conditions, or when one you are watching changes, with the instrument attached.

Signal feed — livesynthetic
  • Lis pendens88 Harrow St, Brooklyn2m ago
  • Deed recorded1402 Sable Ln, Tampa11m ago
  • Tax delinquency9 Bellhouse Ct, Bronx24m ago
  • New mortgage77 Marlowe Blvd, Queens38m ago
  • Entity dissolvedPELL ROW PARTNERS LLC52m ago
  • Code violation304 Pell Row, Brooklyn1h ago
  • Satisfaction filed1220 Kestrel Ave, Queens1h ago
  • Foreclosure filed612 Ordway Dr, Newark2h ago
  • Lis pendens88 Harrow St, Brooklyn2m ago
  • Deed recorded1402 Sable Ln, Tampa11m ago
  • Tax delinquency9 Bellhouse Ct, Bronx24m ago
  • New mortgage77 Marlowe Blvd, Queens38m ago
  • Entity dissolvedPELL ROW PARTNERS LLC52m ago
  • Code violation304 Pell Row, Brooklyn1h ago
  • Satisfaction filed1220 Kestrel Ave, Queens1h ago
  • Foreclosure filed612 Ordway Dr, Newark2h ago
06Worked example
Illustrative walkthrough · synthetic data

A screen that ends in a person, not a pile of shells.

The same synthetic case that runs through this site, seen from the screener.

  1. 01Set the conditionsBrooklyn multifamily · 8+ units · held 5+ years · adjacent parcel under the same resolved controller. Four conditions; each one is a recorded fact, not a model output.
  2. 02Run it resolvedThe screen evaluates against canonical persons, not deed names. VANDERPOOL HOLDINGS LLC and HARROW STREET CAPITAL LLC do not appear as two prospects, they collapse into A. Reyes.
  3. 03Read the why441 Vanderpool St qualifies because the lot next door is held by a different entity with the same resolved controller. The screen shows both deeds, the claim is clickable.
  4. 04Keep it runningSave the screen and it becomes a standing alert: the next acquisition that completes an assemblage pattern reaches you with the new deed attached.
  5. Where you landA shortlist where every row is a person with a reason, and A. Reyes is one row, not six.
Screener — sourced signals1,284 resultssynthetic
Adjacent parcels, same hands3 lots on one block, 2 entities, one shared controllernyc_acris
Free and clearno open mortgage on record since the 2011 satisfactionnyc_acris
Long holdheld 14 years — above the block median of 6nyc_acris
Tax warrant on the ownerstate warrant filed against a controlled entityny_dtf
Open hazardous violation2 class-C violations, no release observednyc_hpd
Every signal is a recorded filing with its source attached. None of them tells you what to do.
07What a list is worth

A list of entities is not a list of owners.

Screen a market and you get back companies. Twelve rows can be one person, which means your outreach is twelve touches on one relationship and your concentration read is wrong by a factor of twelve. Resolution happens before the filter runs, not after you export.

22.0M
properties screened on resolved identity
  1. 01Resolved before filtered

    Conditions evaluate against the canonical person, so “owns more than ten” means ten buildings, not ten shells.

  2. 02Deduplicated on export

    One row per person, with the entities they used listed underneath rather than fragmenting the list.

  3. 03Receipts survive the export

    Every exported row keeps its ids and links back to the records behind it, so the list is still checkable after it leaves us.

  4. 04Every filter is a filed fact

    Held fourteen years, owned free and clear, adjacent lot in the same hands, each traces to an instrument.

Twelve rows, one person, and a concentration number off by twelve.

What actually goes wrong with bought lists
08The signal library

Every filter, and whether it is filed or derived.

The positionWhat the owner holds — the shape of an opportunity
Adjacent parcels, same hands
filed record

Lots that touch, held through different entities, controlled by one resolved person. The clearest statement of intent in the public record, and invisible without resolution.

Free and clear
filed record

No open mortgage on record. Derived from recorded instruments and their satisfactions, not from an estimate of what is left to pay.

Long hold
filed record

Held materially longer than comparable owners nearby. A fact about dates, not a prediction about intent.

Portfolio concentration
filed record

How much of one owner’s position sits in one submarket, asset class, or lender relationship.

The pressureWhat is moving against them — the reason timing matters
Maturing debt
partly derived

Recorded loans approaching term. Where a maturity date is not recorded, it is estimated and labelled estimated, never presented as filed.

Owner-level distress
filed record

Tax warrants, liens and judgments against any entity the owner controls, rolled up to the person rather than trapped at one company.

Building distress
filed record

Open violations, vacate orders, lis pendens and foreclosure filings attached to the asset.

Entity events
filed record

Dissolutions, status changes and agent changes on a controlling company.

What this deliberately is not

There is no 0-to-100 seller-motivation score, and there will not be one. A single number that says act on this is a recommendation dressed as data, it hides which facts drove it, cannot be audited, and quietly transfers your judgement to our weighting.

Model-derived figures do exist in the product for triage. They are badged as estimates and kept off the same surface as cited facts, for the same reason.

09What buyers ask

Straight answers.

Why is there no motivated-seller score? Everyone else has one.

Because a single 0-to-100 number hides which facts drove it, and that is precisely the information you need. Our screen returns the conditions themselves, each with its filing. You can weight them however your strategy demands, a score would quietly replace your judgment with our weighting.

How fresh are the conditions I am filtering on?

Each condition carries the recorded date of the filing behind it, and the coverage page publishes per-source freshness. You are never filtering on an undated assertion.

Can I export the shortlist to my CRM?

Yes. CSV export with the provenance intact. Every row keeps its source and record ids, so the list is still checkable after it leaves Mimir.

What stops the same owner appearing five times under five LLCs?

Resolution runs before the filter does. Conditions evaluate against the canonical person, and the export is deduplicated to one row per person with their entities listed underneath.

Do saved screens cost extra?

No. A saved screen and its standing alert are part of the product, not a tier.

See your market.

Twenty minutes on the live terminal.

See your market