Methodology

How Agentas works

Agentas is TerraNova Capital's own lender intelligence, opened up. Every one of the 2,238 active lenders is compiled from public sources, scored against your deal with the same engine our bankers use, and carries a confidence rating so you can judge it before you rely on it.

Where the data comes from

We compile business information that lenders and their teams make public. We don't buy consumer data, and we don't scrape gated sources.

Lender websites & materials

Public fund pages, strategy descriptions, and disclosed check-size ranges.

LinkedIn & team pages

Deal-team roles — heads of origination, credit partners, structured-finance MDs — not founders or IR.

SEC EDGAR filings

Form D and related filings for fund closes and offering sizes, pulled directly from the SEC.

News & press

Recent deal announcements and fund raises via public news feeds, attached to each fund with a source link.

Contact verification

Deal-team contacts are resolved and email-verified through licensed data providers, with a confidence status per contact.

TerraNova's desk

A working structured-finance team's live lender relationships and closed-financing track record inform which lenders are active and how they're prioritized — public facts only, never a client's private deal terms.

How matches are scored

Each fund gets a 0–100 relevance score — a percentage of the criteria you actually specified. Unspecified filters neither inflate nor penalize a fund.

Deal size fitup to 40 pts

Always counts — how well your amount fits the fund's check-size range. Among funds that bracket your amount, tighter books rank higher: a lender whose range is built around your size outranks a megafund that merely spans it (the megafund stays listed, just lower). Where a fund doesn't disclose a range, we model an estimate from its AUM or comparable lenders: it's marked "~ (est.)" everywhere it appears, its influence is deliberately damped — an estimate never scores as strongly as a sourced range at the same good fit — and it's replaced the moment a sourced or lender-verified figure lands.

Deal typeup to 25 pts

Only when you specify a structure (senior, unitranche, mezz…). Exact match scores full; adjacent types score partial.

Sectorup to 20 pts

Only when you specify a sector. Exact, generalist, and related sectors score on a sliding scale.

Geographyup to 15 pts

Only when you specify a region. Exact, global, and adjacent regions score partial.

Borrower sizeup to 15 pts

Only when you specify an EBITDA band and the fund publishes one. Sparse — a missing band never penalizes a fund.

Bonuses+2 to +12

Small nudges for a known relationship, a contact verified in the last 90 days, an available teaser, or a published fast-close timeline (30 days or less).

Confidence & verification

Every fund carries a research-confidence score of 1–5 and, where we have it, the date it was last verified — both shown on the fund page and result cards. Contacts confirmed through enrichment are marked ✓ Verified with a source; 3,189 deal-team contacts carry a verified email today. If an unlocked email bounces, we re-credit the unlock.

How often it updates

An automated pipeline runs daily — pulling news, checking SEC filings, and extracting facts about fund closes and check sizes. Nothing it finds edits the database directly: every proposed change lands in a “Needs Review” queue and a human approves or rejects it first. That keeps automation from quietly writing something wrong. The freshness date in the hero reflects when the dataset last changed.

The honest fine print

Profiles and contact information are compiled from public sources and TerraNova's own research. They are not provided or endorsed by the funds, and accuracy is not guaranteed. AI-assisted deal parsing and matching are for guidance only — verify outputs before you act on them.

Search the swarm About TerraNova →