Home / Sample briefs / Investors

Investor diligence — sample brief

Anatomy of an investor brief.

An Illustrative end-to-end Veriqa memo on a synthetic quantum target. Same format, same discipline, same verdict vocabulary a real brief uses.

Request a brief on a real target Back to investor diligence
Illustrative — "QubitNova Therapeutics" is a synthetic fictional target. This is not a real assessment, not investment advice, and not an offer or solicitation of any security. Every figure, citation, score, and verdict on this page exists only to demonstrate the format.
Target
QubitNova Therapeutics Illustrative
Brief ID
QN-IV-0001
Date
2026-06-08
Engagement type
Investor — Strategic Brief
Methodology
Rules engine v0.5.0 · Schema v1.0.0

This is a sample. Not investment advice, not a recommendation to buy / sell / hold any security, not an offer or solicitation. "QubitNova Therapeutics" is a synthetic fictional target used to illustrate brief structure only.

01 · Executive verdict

The call. Illustrative

Every Veriqa brief resolves to one of three states — Proceed, Monitor, or Require further diligence. The verdict grades how well the headline claims are supported by the cited evidence, not whether to own anything.

Monitor Sample

Strong scientific narrative; no classical baseline cited on the headline advantage claim; recommend a 60-day evidence-evolution window before deeper diligence.

What this means. The target's underlying science is credible and the team has the credentials they claim. But the load-bearing advantage claim — that their quantum sampler beats classical Monte Carlo on binding-affinity estimation — ships without a named classical solver, a problem size, or a reproduction. Under Veriqa's claim-discipline rules a Proceed verdict is blocked while a core advantage claim is needs-baseline. The verdict caps at Monitor: keep watching, run the brief again when the missing evidence lands, and do not size a position on the headline as written.
02 · Decision frame

The questions this brief answers.

A diligence brief is only as useful as the questions it commits to answering. These are the five the engine grades against on every investor target.

  1. Is the headline advantage claim materially novel?

    Does the work clear the existing public literature, or is the "first" qualified by a narrow setup the framing does not disclose?

  2. Does it beat a named classical baseline at the stated problem size?

    Is there a named classical algorithm, a stated instance size, and a head-to-head result — or is the comparator implicit, missing, or run at a smaller scale than the claim implies?

  3. Is the team credentialed for what they claim?

    Do the founders, scientific advisors, and senior IC have publications, prior shipped systems, or domain operating history that maps to the headline?

  4. Is the IP defensible?

    Is there filed, granted, or assignable IP, and does it cover the load-bearing method rather than peripheral apparatus?

  5. Is the timeline plausible against public roadmaps?

    Does the target's "by year X" claim line up with the public roadmaps of the hardware vendors and standards bodies it depends on?

03 · Score gauges

Three scores, with reasons. Illustrative

Each score is a transparent, rule-based heuristic computed from features of the graded evidence — not a sentiment score, not a statistically validated model. Values are reported out of 10 for at-a-glance reading; the underlying engine carries the full 0–100 trace.

Technical maturity
4.5/10

Capped. Real benchmarks exist on a sub-claim, but the headline advantage claim is needs-baseline — maturity cannot rise above the cap until the baseline is supplied.

Commercial urgency
6.0/10

Moderate. A round-close clock and a published competitor preprint sharpen the timing, but no regulatory deadline or named partner forces a decision this quarter.

Hype / overclaim risk
7.2/10

Elevated. The press release uses "exponential" and "unprecedented" on a result whose preprint scope is narrower than the framing. Superlatives without sources are the dominant driver.

Scores are a transparent rule-based heuristic over the graded evidence — not sentiment, not statistically validated, and not a recommendation. Illustrative

04 · Classical-baseline comparison

Every quantum claim, set against its classical comparator.

The load-bearing test on any quantum claim is the classical baseline at the stated problem size. A row without a named classical solver and a stated size is flagged needs-baseline and caps the verdict at Monitor or below.

Claim Quantum method claimed Classical baseline cited Problem size cited Status
Beats classical Monte Carlo on binding-affinity estimation (headline). Variational quantum sampler on 64-qubit superconducting hardware. None named Not stated Needs-baseline
Reduced sample count vs. classical VMC on a 12-atom test molecule. Quantum variational ansatz, 24 qubits. Open-source VMC implementation, default parameters. 12 atoms, single conformer Baseline cited
Pipeline integration with public protein structure datasets. Classical featurisation; not a quantum advantage claim. N/A — classical only Standard test set Not an advantage claim
"Order-of-magnitude" speedup on drug-screening throughput (press). Hybrid sampler + classical post-processing. Implied, not named Not stated Needs-baseline
Forecasted advantage at 256-qubit scale (roadmap). Future variational sampler. Future classical comparator (TBD). 256 qubits (target) Forecast — ungraded

Two of five rows lack a named classical baseline; under the rules engine this caps maturity and blocks a Proceed verdict. Illustrative

05 · Evidence table

Every sub-claim, sourced. Illustrative

The evidence table is the brief. Each row carries a source ID and a confidence score (0–1) describing how strongly the cited source supports the sub-claim — not a probability that the claim is true.

Sub-claim Source Conf. Notes
Quantum sampler outperforms classical Monte Carlo on binding-affinity estimation. SRC-01 — Press release, 2026-04-22 0.18 Headline claim. No benchmark, no classical baseline named, no problem size stated. Flagged needs-baseline.
Variational ansatz reduces sample count on a 12-atom test molecule vs. open-source VMC. SRC-02 — arXiv:2604.01234 (preprint, 2026-04-08) 0.72 Pre-print. Classical comparator named, problem size stated, code released. Not yet peer-reviewed; no independent reproduction.
Method for sampler-classical hybrid binding-affinity estimation. SRC-03 — USPTO patent application US 2026/0123456 A1 0.65 Application, not granted. Covers the headline method; assignment to QubitNova Therapeutics confirmed in filing.
Post-quantum cryptographic posture on customer data. SRC-04 — NIST FIPS 203 / 204 / 205, 2024 0.90 Standards body. Target's data-handling claim maps cleanly to FIPS 203 KEM and FIPS 204 signatures; migration plan dated.
Hardware availability for 64-qubit superconducting runs. SRC-05 — Public hardware-vendor roadmap PDF, 2026 0.60 Vendor roadmap supports availability of the qubit count claimed for the headline run, but timing tolerance is wide.
Lead investigator publication record in quantum chemistry sampling. SRC-06 — Public author listing on arXiv (8 first-author preprints, 2021–2026) 0.82 Credentialed in the subfield. Recent record consistent with the methods described in SRC-02.
"Order-of-magnitude" drug-screening speedup. SRC-07 — Investor deck slide 14 (NDA) 0.22 No benchmark, no comparator, no instance size. Superlative framing exceeds preprint scope. Flagged needs-baseline.
Forecasted advantage at 256-qubit scale by 2028. SRC-08 — Company roadmap, undated 0.30 Forecast; not graded. Hardware availability at that qubit count by 2028 is contingent on SRC-05 vendor execution.

Confidence (0–1) measures how strongly the cited source supports the sub-claim, as graded by the engine — not a probability the claim is true. Source numbering is stable across revisions of this brief. Illustrative

06 · Missing evidence

What would change the call.

Each item below is a specific deliverable that, if supplied and graded, could lift the verdict. The list is intentionally narrow: it is what the next 60–90 days of diligence should chase.

07 · Validation plan

The next 90 days, in five steps.

A monitor verdict is only useful with a plan attached. These are the steps and deliverables that would let Veriqa re-grade the brief and either lift the verdict to Proceed or move it to Require further diligence.

  1. 1

    Day 0–10 · Source request

    Formal request to QubitNova for the named classical solver, version, problem instance, and seed used in the headline run. Deliverable: a written response or a refusal on the record.

  2. 2

    Day 10–30 · Independent re-run of SRC-02

    Engage an external chemistry-aware engineer to re-run the released code on the same 12-atom instance. Deliverable: a one-page reproduction memo with sample-count and wall-clock results.

  3. 3

    Day 30–50 · Hardware audit

    Confirm with the named hardware vendor that the 64-qubit run in SRC-01 was scheduled and executed at the claimed scale. Deliverable: vendor-confirmed run log or a flag that no record exists.

  4. 4

    Day 50–75 · IP and team check

    Pull USPTO PAIR status on SRC-03, confirm inventor assignments, and validate scientific advisor commitments beyond the deck. Deliverable: an IP-and-team appendix to this brief.

  5. 5

    Day 75–90 · Re-grade

    Veriqa re-runs the brief with the new evidence. Deliverable: a versioned successor (QN-IV-0002) with a fresh verdict, a diff against this brief, and an updated missing-evidence list.

08 · Reviewer status

How this draft cleared the gate.

Held in draft until reviewer approves. A Veriqa brief cannot be delivered until an internal reviewer approves it — the software enforces the gate, the brief cannot be released around it. Reviewer credential standard: graduate degree in quantum information, physics, or applied cryptography, or equivalent hands-on grading experience.

We do not yet represent that an independent outside expert reviews every report. For high-stakes decisions we recommend independent expert review on your side in addition to ours.

Methodology version. Graded under Veriqa rules engine v0.5.0, report schema v1.0.0, 2026-06-08. Match this against the changelog to reproduce. The same rules engine runs whether the report was produced offline or with AI enrichment; AI enrichment cannot lift a needs-baseline flag, uncap a maturity score, or unlock a Proceed verdict.

Decision-support and general information only — not investment, financial, legal, or tax advice, not a recommendation to buy, sell, or hold any security, and not an offer or solicitation. "QubitNova Therapeutics" is a synthetic fictional target; every figure, citation, score, and verdict on this page is Illustrative.

Bring us a real target.

You've seen the format. Send us the company, the claim, or the thesis on your desk and we'll turn it into a graded, source-linked brief your IC or board can stand behind.

Request a brief on your real target Back to investor diligence

Decision-support and general information only — not investment, financial, legal, or tax advice, not a recommendation to buy, sell, or hold any security, and not an offer or solicitation of any security.