Lensential Venture

SAME CLAIMS, TWO ANSWERS

The assistant argues; Lensential Venture grades.

We gave the specimen company's claims to a general-purpose chat assistant with one prompt: "Here is a startup we are looking at. Should we invest? Write me a one-page memo." On the left, passages from its answer. On the right, the Lensential Venture screen on the same claims, checked against the documents behind them.

A CHAT ASSISTANT

"Proceed to a term sheet, subject to confirmatory diligence."

LENSENTIAL VENTURE · THE SCREENSPECIMEN

Overall B-

A screening judgment from limited evidence, not an investment recommendation. The decision stays yours.

A CHAT ASSISTANT

"Thornlark's microfluidic cartridge and bench reader return a result in eighteen minutes from a finger-stick sample, run by a nurse in the exam room."

LENSENTIAL VENTURE · THE SCREENSPECIMEN

UNVERIFIED

18-minute sample-to-result turnaround.

Company bench measurement. No third-party run, no published protocol, no independent replication.

A CHAT ASSISTANT

"Commercial validation is early but real: four signed evaluation agreements, one paid placement, and $240,000 ARR before FDA clearance."

"Traction is thin: $240,000 ARR rests largely on one paid placement, and evaluation agreements often fail to convert."

LENSENTIAL VENTURE · THE SCREENSPECIMEN

REFUTED

USD 240,000 ARR.

The underlying invoice is USD 18,000 to place the product once, so it is not recurring revenue.

Both name risks; only Lensential Venture says which claims it checked, against what, and which it could not.

Read the assistant's full answer

Investment Memo: Thornlark Diagnostics

Recommendation

Proceed to a term sheet, subject to confirmatory diligence. Thornlark addresses a clear clinical pain point with existing reimbursement and a credible team. The main risks are regulatory timing and whether pilot interest converts into durable revenue.

The Opportunity

Urgent-care clinics currently send a high-volume analyte to an outside lab and wait about four hours. Thornlark's microfluidic cartridge and bench reader return a result in eighteen minutes from a finger-stick sample, run by a nurse in the exam room. The analyte already carries a paying CPT code, so clinics bill as they do today and adoption does not hinge on new reimbursement. The company sizes the market at $4.1 billion across roughly 8,900 US urgent-care sites; neither incumbent analyzer reports in under twenty minutes.

Strengths

Thornlark owns the chemistry and cartridge, has filed a provisional patent, and reads results with an on-device machine-learning classifier. Commercial validation is early but real: four signed evaluation agreements, one paid placement, and $240,000 ARR before FDA clearance. The CTO has eleven peer-reviewed papers in assay chemistry; the commercial lead grew a diagnostics line to $30 million in revenue. Three strategic acquirers have paid more than 6x revenue for point-of-care diagnostics companies in the last four years.

Key Risks

The 510(k) submission is not planned until Q3 2027, so clearance likely lands in 2028. The $2.5 million must bridge that gap, and any delay could force a dilutive raise. Traction is thin: $240,000 ARR rests largely on one paid placement, and evaluation agreements often fail to convert. A provisional filing is not an issued patent, and freedom to operate should be checked. The incumbents could also respond with a faster cartridge on readers clinics already own.

Deal Terms

The round is $2.5 million at a $14 million post-money, with $900,000 soft-circled. The 2025 $750,000 SAFE converts at a $9 million cap, adding dilution beyond the headline. The implied pre-money of roughly $11.5 million is reasonable for a diagnostics company with early revenue and a strong team, though not cheap pre-clearance.

Next Steps

Request clinical performance data (sensitivity and specificity), confirm the regulatory strategy with an FDA consultant, speak with the evaluation sites, and verify that runway covers the submission.

The assistant's answer is unedited. Thornlark Diagnostics is a fictional company.

THE REPORTS

Seven questions, in the order a deal travels.

Pick a question. Each opens the reports that answer it, what you send and what you get, and the agents that do the work.

TRIAGE AND SCREEN

PROFILE

VERIFY

LAND

DECIDE AND SHARE

Which of these deserve our time?

Funnel triage · The Screen, with an exit read

Funnel triage sorts a whole cycle of deal flow against your thesis in one pass, and says in writing why each company failed. The Screen tells you what one company actually builds, what is special about it and what twenty other companies are also doing. It grades six categories, plus the deal terms when known, on evidence written down before any grade, and sets the round's price against what buyers have paid for similar companies.

SPECIMEN

THE SCREEN · BRIEFSPECIMEN

Thornlark Diagnostics

Thornlark Diagnostics, Inc., a Delaware corporation incorporated March 14, 2024, Austin, Texas. Eleven full-time employees.

USD 2,500,000 seed at a USD 14,000,000 post-money valuation. No lead identified. A USD 750,000 SAFE from 2025 converts at a USD 9,000,000 cap, which takes roughly 5.4% off the founders before new money arrives.

Screened August 8, 2026 · Overall B-

Category	Grade	Rationale
Team	B	The CTO has eleven peer-reviewed papers in assay chemistry, and the commercial lead carried a diagnostics line to USD 30 million. But no one on the team is named on an FDA-cleared 510(k), and the plan needs that clearance in nineteen months.
Market	B	Roughly 8,900 US urgent-care sites and a reimbursable billing code that already pays, but the deck's USD 4.1 billion market counts the whole panel, and Thornlark addresses one analyte on it.
Competition	C+	Two incumbents own the shelf and a third filed a competing clearance application in April, but none of the three reports a result in under twenty minutes, the only axis on which Thornlark is ahead.
Differentiation	B-	The microfluidic cartridge is genuinely novel and the provisional patent filing is real, but the classifier is trained on 1,240 samples from a single site.
Metrics	C	Four signed evaluation agreements and one paid pilot, but the pilot is USD 18,000 of instrument placement, not test revenue, and the deck's USD 240,000 ARR is a run-rate built from it.
Exit potential	B	Three strategic buyers paid more than 6x revenue for point-of-care assets in the last four years, but every one of those targets already had clearance.
Deal terms	C+	A standard seed at a defensible price for the stage, but the 2025 SAFE cap sits 36% below a round no lead has priced.

Screening judgment from limited evidence. Not diligence, and not an investment recommendation. Evidence is written and labeled before any grade is given; a category whose evidence is entirely company-stated is capped at B-.

What can we establish before we call the founder?

The Profile

The overview before deep diligence: eighteen sections built from public sources, such as registries and filings, patents, web archives, and the company's own site and product. Deck claims are tested, never assumed. Every material claim goes into a claim check with one of the six labels, and every claim we cannot verify becomes a question, not an accusation.

SPECIMEN

Claim check

THORNLARK DIAGNOSTICS · SPECIMEN

VERIFIED

Four signed evaluation agreements

Executed copies in the data room, four distinct counterparties. None carries a purchase commitment.

PARTIAL

USD 4.1 billion addressable market

The site count and the payment amount are both verified. The payment covers the whole test panel; the product measures one analyte.

UNVERIFIED

18-minute sample-to-result turnaround

Company bench measurement. No third-party run, no published protocol, no independent replication.

REFUTED

USD 240,000 ARR

The underlying invoice is USD 18,000 to place the product once, so it is not recurring revenue.

REPORTED

Classifier trained on 1,240 samples

Company-stated, consistent across the deck and the technical memo. No data-use agreement was provided.

MISSING

Regulatory clearance history for the team

No team member is named on an FDA-cleared 510(k). Searched the FDA premarket database under every name variant.

Are the documents and the numbers real?

Diligence report, numbers scope

Four agents, always in this order. Who made each file, when, and with what software. The company's timeline, from records it does not control. Whether the actual financials record real activity or were generated by the forecast. Whether each revenue line started after the license, launch or product it depends on. We open every class of document in the data room, bank statements included.

SPECIMEN

FINDINGS REGISTER · THE NUMBERSSPECIMEN

Ref	Finding	Direction	Firmness	The company's best answer	Does it survive?
F-07	USD 240,000 ARR is one USD 18,000 invoice, grossed up and multiplied by twelve	▼ ADVERSE	VERIFIED	"Every early company annualizes a pilot."	Yes, in full. It is the company's arithmetic on the company's documents.
F-06	One customer credit, USD 18,000, in 31 consecutive monthly bank statements	▼ ADVERSE	VERIFIED	"We are pre-commercial."	Yes. The answer concedes it.
F-10	Recomputed runway after the round is 11.8 months, against 18 in the model	▼ ADVERSE	PARTIAL	"The study can be scoped down and hiring slowed."	Partly. Both levers are real, and neither is quantified in the file.
F-15	The bank record is complete, consecutive and free of gaps	▲ FAVORABLE	VERIFIED	None needed.	Yes. Every credit in the company's life is identifiable to a counterparty.

Four of forty findings in the specimen diligence report, which ranks every finding by how well it survives the company's best answer.

Is the product real, and who else could sell it?

Competitive intelligence, with a white-space map · Diligence report, technology scope

We test the live product from outside, with no accounts created and no terms accepted. We price what exists and what the documents describe, in two columns that never mix. Then we map who else could sell it, and say whether the company sells a product, a component of a product, or a technology in search of one.

SPECIMEN

The US field

SPECIMEN

Who sells what, graded cell by cell, as of August 8, 2026

	What they sell					How it is delivered		Who buys
	This analyte on the menu	Point-of-care reader	Single-use cartridge	Result under 20 minutes	On-device classification	US clearance	Waived format	Urgent care	Retail clinics
Thornlark Diagnostics (the company)	●	●	●	◐	◐	○	○	●	○
Brackwell Point	○	●	●	○	·	●	●	●	●
Quillfern Dx	·	●	●	·	·	●	●	●	●
Tamsworth Rapid	·	●	●	·	·	●	●	●	●
Penhallow Medical	·	●	●	·	◐	●	●	·	●
Merrowbank Dx	●	●	·	·	·	●	·	●	·
Aldmoor Systems	●	·	·	·	·	●	·	·	·
Ashgrove Microfluidics	◐	·	●	◐	·	·	·	·	·
Larkspire Systems	·	·	●	◐	◐	○	·	·	·
Harnwick Analytics	●	·	·	·	·	·	·	●	·
Vendors with evidence, of 14 counted	5	6	6	0	0	8	5	7	5

● Evidenced · ◐ Claimed only · ○ Planned · · Not found, with where we looked

Ten of the sixteen rows and nine of the 22 columns on the specimen's US chart. The full paper grades 22 vendors in 484 cells, every cell with its source.

Who are we dealing with?

Diligence report, people and round scope · Alignment paperBETA

We check the team's stated credentials against primary records and report whether each error overstates or understates them. We check every named investor against that investor's own records and total the round. When an insider anchors the round, we rebuild their position from the signed documents and ask whether their reason to invest applies to you.

SPECIMEN

FINDINGS REGISTER · PEOPLE AND THE ROUNDSPECIMEN

Ref	Finding	Direction	Firmness	The company's best answer	Does it survive?
F-01	The 2025 SAFE is a post-money instrument converted on a pre-money basis in the company's cap table	▼ ADVERSE	VERIFIED	"The cap table is illustrative."	Yes, in full. The answer concedes the arithmetic, and nothing shows the investor has agreed.
F-04	No Section 83(b) election on file for one of the two founders	▼ ADVERSE	MISSING	"It was filed; the copy is with the founder's accountant."	Weakly, until the receipt is produced. It matters because the election cannot be fixed later.
F-11	Eleven stated full-time employees reconcile to six W-2 employees	▼ ADVERSE	VERIFIED	"Contractors work full time."	Narrowly. Contractor cost is not in the per-head figures the model uses to plan the team.
F-37	The commercial lead has carried a diagnostics line at scale	▲ FAVORABLE	PARTIAL	None needed.	Yes. Two independent sources corroborate it, at the level of the product line.

Four of forty findings in the specimen diligence report.

Can we make money at this price?

The Exit Brief and the live exit workbook · The Note Brief and the Return Grid

Investors ask this first. We carry your stake through Series A, B and C, with option pool refreshes, SAFEs, notes and preference stacks, and grade every exit against what buyers actually paid for comparable companies. One page gives the answer; the workbook shows the math.

SPECIMEN

EXIT BRIEFSPECIMEN

Thornlark Diagnostics

At $14M post-money, returning 3x after the Series B needs a $73.5M exit, inside what buyers have paid. That takes about $10M of revenue at their median multiple, against $18K today. The return depends on regulatory clearance.

$14MENTRY VALUATION

15.9xRETURN AT $390M, THE HIGHEST PRICE PAID, AFTER THE SERIES B

$73.5MEXIT NEEDED FOR 3X, AFTER THE SERIES B

19%OF THE HIGHEST PRICE PAID

Terms: $1.25M for 8.9% of the company, fully diluted after each round the plan raises (Series A and B), pro rata.

Likely exits, and what they return

Exit value	Support	Multiple after Series B
$40M	EVIDENCED	1.63x
$91M	EVIDENCED	3.71x
$180M	EVIDENCED	7.34x
$390M	EVIDENCED	15.9x
$650M	HYPOTHETICAL	26.5x

Evidenced at or below the highest on-point comparable ($390M, Luxtra Diagnostics); Hypothetical above it.

What each target multiple needs

Target	Exit after Series B	Support	Revenue at 7.67x
3x	$73.5M	EVIDENCED	$9.59M
5x	$122.5M	EVIDENCED	$16M
10x	$245.1M	EVIDENCED	$32M

What buyers paid

Target	Buyer	Year	Value	Multiple
Pellwyn Point-of-Care	Harrowgate Instruments	2023	$189.1M	6.10x
Corvane Rapid Assays	Ostrevale Health	2024	$145M	6.59x
Luxtra Diagnostics	Aldersgate Life Sciences	2022	$390M	9.40x

Median 7.67x across four comparables, one of them a reference for scale. Screening arithmetic from the workbook built on the same file. Not an investment recommendation.

Can we defend the decision, and what do we tell others?

Diligence report · Founder questions and document request · Committee decision paper · Investment memo

The diligence report ranks every finding by how well it survives the company's best answer. The founder questions turn findings into neutral questions and end with the three to five documents that would settle them. The committee paper puts the case for and the case against at their strongest, and names the one event they disagree about. The memo goes to co-investors and LPs, with the evidence and the partners' judgment kept apart.

SPECIMEN

QUESTIONS FOR THE COMPANYSPECIMEN

Q-01
Will you run ten unselected consecutive patient samples on a reader at a site we nominate, with our observer present and the acceptance protocol written down beforehand?
From: finding F-23
Resolves: the only differentiating claim in the file, and the 22.6% cartridge fault rate the run log implies.
Q-04
Has the 2025 SAFE investor agreed to convert on the basis shown in your cap table, and does any further convertible instrument, note or side arrangement exist that could trigger its most-favored-nation clause?
From: finding F-01
Resolves: about 145 basis points of the post-close table, and whether the cap table circulated to investors is the table that will close.
Q-06
Under what agreement were the 1,240 training samples obtained, and does it permit commercial use of a model trained on them?
From: finding F-25
Resolves: whether the classifier is an asset the company can sell.
Q-10
Is the USD 240,000 ARR figure being withdrawn from the deck, and what figure replaces it?
From: finding F-07
Resolves: how the company handles a number of its own that does not survive examination.

Four of the specimen's ten priority-one questions. Each cites the finding behind it and says what an answer would settle.

TRIAGE AND SCREENWhich of these deserve our time?

PROFILEWhat can we establish before we call the founder?

VERIFYAre the documents and the numbers real?

VERIFYIs the product real, and who else could sell it?

VERIFYWho are we dealing with?

LANDCan we make money at this price?

DECIDE AND SHARECan we defend the decision, and what do we tell others?

HOW WE DELIVER

Finished files, signed by a principal and presented by the analyst who ran them.

Signed by a principal.

Every paper is read, challenged and signed by a principal before it reaches you. Delivery times include that review and start when your input arrives. You sign the papers sent in your own name, such as a letter to your investors.

Formats follow the reader.

Reports come as PDF, because evidence should not be edited afterward. Workbooks come as live Excel, so you can run them again. The memo and the founder questions come as editable Word.

Presented to your partners.

An analyst walks your partners or your committee through the findings, on a call or in the room.

In your house style, or plain.

Your mark, typeface and credit line are applied when the report is rendered. Plain is a tested default, equal to house style.

Corrected in the open.

If a finding changes after the founder answers, you receive the corrections register, with the earlier reading kept.

Start with what you have.

A name or a website is enough for a screen or a profile. A deck enables the claim check. The system works best with a full data room; it has been tested on data rooms of more than 200 files.

How you pay.

Per paper; a screening retainer of five or ten screens a month, each within 24 hours; or the full series when you decide to pursue a deal. Fees are agreed before work begins.

ONE STANDARD

Six rules every agent follows.

1
Label every finding twice.
Which way it points, for or against the company's account, and how firm it is.
2
Test the founder's best answer, not just the claim.
Findings are ranked by how well they survive it.
3
Never conclude intent.
"No independently verifiable financial information" is defensible; "fabricated" is not.
4
Not finding evidence does not prove something did not happen.
The caveat sits beside the finding, every time.
5
Keep a corrections register.
We publish it for every version that circulated.
6
Show how to reproduce it.
If a third party cannot follow the steps, it is not a finding.

VERIFIED

Established against an authoritative or independent source, and reproducible.

PARTIAL

Established in substance, with a material qualification.

UNVERIFIED

Asserted, and neither confirmed nor contradicted. Not a false claim.

REFUTED

Contradicted by an authoritative source or the company's own materials.

REPORTED

A company figure carried for analysis, with no independent support.

MISSING

Not disclosed and not found. We name where we looked.

Rule 5, in practice.

CORRECTIONS REGISTERSPECIMEN

Six entries. Three of our errors favored the company and three were harsh to it. Nothing is deleted.

Ref	As reported	Corrected	Our error ran
C-01	The interim memo of August 7, 2026: the laboratory's certificate bears an effective date before the subsidiary was formed.	Withdrawn in full. The effective date is February 17, 2026, fifteen days after formation; the earlier reading took the application date for it.	Harsh to the company
C-02	The Screen of August 8, 2026: the 2025 SAFE cap takes roughly 5.4% off the founders.	The instrument is post-money. It converts to 6.85% of the post-close table.	Generous to the company
C-03	The Screen and the Profile of August 8, 2026: four signed evaluation agreements, VERIFIED.	Amended to PARTIAL. Three are signed by both parties; the fourth carries only Thornlark's signature.	Generous to the company

From the specimen diligence report. We publish the register for every version that circulated.

DEEP DILIGENCE

Nine agents. One question each.

These are the questions a partner asks once a company is worth real time. Each agent answers one, in a fixed order, and labels what it finds.

OPENDocument historyWho made these documents, when, and how often were they revised?

We read every file's metadata before anyone opens its contents. The comparison between files is the finding, not any single value.

Catches: files that describe a long history but were created days before the data room opened.

DATEPublic-record timelineDid the company really exist for three years, or is the story new?

We date the company from registers, licenses, domain records and archived copies of its website, and show how its description of itself changed over time.

Catches: a pivot told as a long history, and revenue reported before the license that permits it.

TESTFinancial model forensicsAre the actual financials really actual?

We apply the forecast's own growth rates to the reported history, month by month, and score how often they predict the next month exactly.

Catches: history that follows the spreadsheet's own growth assumptions and changes rate on a calendar date.

TESTRevenue and tractionIs the revenue real, and where does it come from?

We break revenue down by line and month and test whether each line started after what it depends on. We convert churn to annual retention, rebuild LTV from the company's own inputs and check for acquisition cost in the months customers arrived.

Catches: subscription revenue booked before the public launch.

TESTProduct from the outsideDoes the live product do what the deck says?

We test the product with no accounts created, no terms accepted and no forms submitted. Untestable is recorded as untestable, with no adverse inference.

Catches: "proprietary data" whose field names match an open government dataset.

TESTBuild and replication costWhat would it take to build what the founder describes, and to rebuild what exists?

Two columns that never mix: what exists, priced as the simplest build that matches what we observed, and what the documents describe, priced against comparable teams. Ranges with stated assumptions, never a quotation.

Catches: a "model" that behaves like a lookup table.

TESTCompetitive intelligenceWho else could sell this, and is there white space?

We define the field by how the product works, not by its category, and grade every vendor against one set of columns.

Catches: a company selling one part of a product at the price of the whole.

PEOPLEFounder credentialsDo the team's credentials hold up?

Each claim is checked against primary records: the person's own profiles, the employer's own site, the program's own list. Where no public record exists, as is often the case for degrees, we say so, never treat the gap as proof, and put the claim to the founder in writing.

Catches: a title one level above the real one, and a founding role that began after the founding.

PEOPLEInvestors in the roundWho is really in the round?

Every named investor is checked against its own website, never press or aggregators. We separate investments from accelerator programs, then total the amounts against the claimed round.

Catches: an institutional investor that turns out to be the institution's alumni angel network.

When an insider anchors the round, alignment analysis rebuilds their position from the signed documents and asks whether their reason to invest applies to you. It is in beta after three live runs, and so far it ends in a document request more often than a verdict. BETA

COMPETITIVE INTELLIGENCE, WITH A WHITE-SPACE MAP

We grade the whole field, cell by cell.

In white-space mode, every vendor in the field is graded against one set of columns in four blocks: what they sell, how it is delivered, who buys, and where. Every cell gets a grade and a source. A column with few vendors shows room, not proof that anyone wants it, so we test every gap four ways before we call it white space.

1
Wanted in writing? Is the need written down in a buyer's own document?
2
Fenced? Does a rule or an approval keep vendors out, or let them in?
3
Free or bundled elsewhere? Does a workaround cost the buyer nothing?
4
Held by someone else? Patents, exclusive assets, channels.

See the US field specimen

WHITE SPACE

Thin in every chart, wanted in writing, and nothing free or held closes it.

CONTESTED WHITE SPACE

Thin and wanted in writing, but a free workaround, a fence or a patent stands in the way.

EMPTY FOR A REASON

Thin, but nobody has written the need down, or it is free elsewhere, fenced against the company, or held by someone else.

NOT WHITE SPACE

Thin only in one chart or one column. Counted across the whole field, the space is taken.

REGISTERSPECIMEN

Gap	In the field	Reading
A result in under 20 minutes	US field: 0 of 14 evidenced, 2 claimed only. Outside the US: 1 of 7 evidenced.	CONTESTED WHITE SPACE
On-device classification	US field: 0 of 14 evidenced, 2 claimed only. Outside the US: 0 of 7.	EMPTY FOR A REASON
Selling in the European Union	US field: 1 of 14 evidenced. Outside the US: 5 of 7.	NOT WHITE SPACE
Retail clinics, US South	0 of the 5 vendors that sell the test.	EMPTY FOR A REASON

Gap 1 · A result in under 20 minutes

SPECIMEN

Wanted in writing?: Yes. The specialty society's 2026 practice statement names turnaround as a barrier: 34% of eligible patients leave before the result returns.
Fenced?: Yes, both ways. US clearance and a waiver keep out the one vendor that shows a result this fast, because it sells only under a European mark. A cleared competitor filed in April 2026.
Free or bundled elsewhere?: Partly. A courier send-out costs a site nothing at the point of use, and 61% of sites use one. It takes four hours, not 18 minutes.
Held by someone else?: Partly. A competitor's pending patent application claims the range the cartridge sits in. It is granted nowhere yet.

Reading: CONTESTED WHITE SPACE. The buyer has written the need down, and a free workaround and a cleared fast follower stand in the way.

"Is there white space? One gap is real, and it is contested. No US vendor shows a result for this analyte in under 20 minutes, and the buyer has written that need down."

The short answer of the specimen paper, Thornlark Diagnostics

White-space mode needs a company at the center of the field and an agreed budget, because checking each cell costs about one and a half times the research. It has run on one real engagement so far.

THE EXIT MODEL

Can we make money at this price?

We carry your stake through Series A, B and C and grade every exit against what buyers actually paid for comparable companies. The floor comes from the capital the company still needs and from keeping the founders motivated: combined founder ownership below about 20% at the last round is flagged, because it threatens the investment itself.

What the model handles

Post-money percentage SAFEs
Most-favored-nation clauses, read broadly until the SAFE itself is read
Convertible notes with interest
Option pool refreshes
Liquidation preference stacks
Change-of-control repayment
Follow-ons, judged on the new money first

What it refuses, rather than guess

Participation caps
Ratchets
Pre-money SAFEs

Every exit carries one of three bands, set by the engine: EVIDENCED, at or below a price buyers have paid; STRATEGIC, only with a named precedent; HYPOTHETICAL, above anything paid.

See the Exit Brief specimen

Holding a SAFE or a note? The Note Brief puts what it pays on one page.

NOTE BRIEF · WHAT YOUR CONVERTIBLE NOTE RETURNSSPECIMEN

Thornlark Diagnostics

Your cap decides your price. The round the company plans is priced at $20M pre-money, and 20% off that is $16M, above your $12M cap, so you convert at the cap. On a sale before the next round, you are repaid in full, $392,000 with interest, once the sale covers the $1.34M owed on all the notes. If the sale brings more than your $12M cap, your converted share is worth more than that repayment. The 8% interest raises what you are owed but does not move the $12M point.

$350,000YOUR CHECK

$12MYOUR VALUATION CAP

3.27%YOUR SHARE IF IT CONVERTED TODAY

2.29%YOUR SHARE AFTER THE NEXT ROUND

REPAID IN PARTbelow $1.34M

REPAID IN FULL WITH INTEREST$1.34M to $12M

YOUR SHARE IS WORTH MOREabove $12M

On a sale before the next round, converting pays more than repayment once the sale price passes the cap.

WHAT IT DOES NOT DO

Our limits.

We do not recommend for or against an investment. A grade is investor judgment on the evidence, not a prediction of returns.
We give no legal, tax or IP advice. Each question goes to your counsel or adviser with the date the answer is needed.
Portfolio monitoring, term-sheet review, reference calls and a valuation frame are deliberately not built into Lensential Venture. Each will be written only after we have done that work for real.
The exit model ends at Series C, and it refuses participation caps, ratchets and pre-money SAFEs rather than guess.
Not finding a public record does not make a claim false. We say where we looked.
We do not run background checks of the kind licensed screening firms run: no criminal, credit or litigation searches.
White-space mapping needs a company at the center of the field.

WHERE THOSE QUESTIONS GO

Area	The question	Routed to	Needed by
Intellectual property	A competitor's pending patent application claims the range the cartridge sits in. Would the cartridge have freedom to operate if it is granted?	Your IP counsel	Before the term sheet
Legal	The cap table converts the post-money 2025 SAFE on a pre-money basis. Does anything bind the investor to that basis, and what can its most-favored-nation clause reach?	Your counsel	Before signing
Tax	Does the fund's stake qualify as qualified small business stock?	Your tax adviser	Before the wire

Three items from the specimen's questions register. We flag each question and route it; the advice is your adviser's. Thornlark Diagnostics is a fictional company.

THE STAGE ENGINE

Seven stages. Skip any, never change the order.

The engine proposes a route and confirms it with you before the heavy work starts. It keeps one engagement file across sessions and stops at the end of each stage to record what changed.

Document provenanceRegistry chronologyData room forensicsTraction verificationProduct verificationBuild assessmentCompetitive positionFounder verificationCap table verification

Inside Verify the order is fixed: open the files, date the company, test the numbers and the product, then look at the people.

Every diligence engagement ends with a question list and a document request.

ALL 19 AGENTS

The agents behind the reports.

Stage	Agent	What it does
Triage	Funnel triage	Sorts a whole cycle of deal flow against your thesis, and says why each company failed.
Screen	Deal screen	Grades a company on six categories, evidence first, and sets its price against comparable exits.
Profile	Company profile	Eighteen sections from public sources, with the deck checked claim by claim.
Verify	Document provenance	Reads every file's metadata before anyone opens its contents.
Verify	Registry chronology	Dates the company from records it does not control.
Verify	Data room forensics	Tests whether historical figures record real activity or came from the forecast.
Verify	Traction verification	Checks each revenue line against what it depends on, month by month.
Verify	Product verification	Tests the live product from outside, with no special access.
Verify	Build assessment	Prices what exists and what the documents describe, in two columns.
Verify	Competitive position	Maps the field by how the product works, with an optional white-space map.
Verify	Founder verification	Checks credentials against primary records only.
Verify	Cap table verification	Checks every named investor against its own records.
Land	Diligence report	Ranks findings by how well they survive the company's best answer.
Land	Founder questions	Turns findings into neutral questions and a short document request.
Land	Exit model	Carries the stake through Series C and grades every exit against what buyers paid.
Land	Note return grid	Prices a SAFE or a note against the next round and the sale.
Land	Alignment analysis BETA	Asks whether an insider's reason to invest applies to you.
Decide	Committee decision paper	Puts both cases at their strongest and names the event they disagree about.
Share	Investment memo	Keeps the evidence and the partners' judgment apart, for co-investors and LPs.

The stage engine runs them in order and keeps the engagement file.

RAISING A SERIES A OR B?

Run Lensential Venture on your own company before your investors do.

See raise readiness

EXPLORE

Lensential PE Lensential Consult Lensential Scout PREVIEW Build with us

Send us the last deal you passed on, or the last one you funded.

We screen it free within 24 hours, once per firm. Or ask for a specimen report first.

Request a sample Book a call

Is what you were told true?

Investment Memo: Thornlark Diagnostics

Recommendation

The Opportunity

Strengths

Key Risks

Deal Terms

Next Steps

Which of these deserve our time?

Thornlark Diagnostics

What can we establish before we call the founder?

Claim check

Are the documents and the numbers real?

Is the product real, and who else could sell it?

The US field

Who are we dealing with?

Can we make money at this price?

Thornlark Diagnostics

Can we defend the decision, and what do we tell others?

Rule 5, in practice.

Holding a SAFE or a note? The Note Brief puts what it pays on one page.

Thornlark Diagnostics

Run Lensential Venture on your own company before your investors do.

Send us the last deal you passed on, or the last one you funded.