Speedup without a baseline
"Exponential speedup" is asserted without a named classical baseline — no FEP, DFT, or MD comparator at a stated problem size. With nothing to measure against, the claim is unfalsifiable, not validated.
Home / Solutions / Pharma & R&D
Enterprise pilot
Quantum drug-discovery claims,
separated from the wet-lab reality.
Veriqa's Molecular Design Agent separates AI-assisted design, classical simulation, and selective quantum refinement — then produces a conservative, patent-ready evidence package with the gaps flagged. It grades the evidence behind a claim. It does not run quantum hardware or claim quantum advantage on your behalf.
01 · The credibility gap
Where quantum-for-discovery goes wrong.
The science is real, but the marketing has outrun it. Four failure modes recur in decks, filings, and vendor pitches — each one a liability for the team or board that acts on it.
AI/ML dressed as quantum
Wins from AI/ML-driven design or DFT/ML surrogates get rebranded as quantum computation. The result may be genuine — but attributing it to a quantum step that did no work is an overclaim.
No wet-lab behind it
In-silico binding scores are presented as outcomes with no measured affinity, no assay, no validation cohort. A predicted number is a hypothesis, not a hit.
IP risk from overclaiming
Inflated language in patent filings, grant applications, and investor decks can undermine prosecution, diligence, and credibility later — when the gap between claim and evidence is examined under scrutiny.
02 · The workflow
What the agent does.
Veriqa ingests a design or claim package and pulls it apart by method. Every advantage claim is forced to carry a classical comparator; every missing experiment is named. The output is a conservative evidence log and a verdict — not a headline.
- 1
Ingest the package
Design dossier, claim, deck, draft filing, or simulation output — with sources attached.
- 2
Separate by method
Map AI design vs. classical simulation vs. quantum-refined steps. No conflation.
- 3
Demand a comparator
Any advantage claim must carry a named classical baseline (FEP / DFT) at a stated problem size.
- 4
Flag the gaps
Surface missing wet-lab, affinity, or assay data and what it would take to close each gap.
- 5
Log + verdict
Conservative evidence log with source IDs and confidence, plus a verdict and validation plan.
03 · Conservative by construction
Restraint is the differentiator.
This domain is full of overclaiming. Our edge is the refusal to record anything we cannot substantiate.
No proven quantum advantage is recorded without a named benchmark, a stated problem size, and a classical baseline (for example FEP or DFT). Any advantage claim missing one of these is flagged needs-baseline — it cannot raise the maturity verdict, and the agent says so plainly. The discipline is enforced in the workflow, not just promised in copy. It exists to protect your IP and your credibility when the evidence is examined later.
Decision-support only. Veriqa's grading and verdicts assist your team's judgement; they are not a substitute for independent scientific, regulatory, legal, or investment review, and require your own verification.
04 · What you get
A defensible evidence package.
Every engagement returns the same structured artifacts — built to survive diligence, prosecution, and a skeptical board.
- Claim / method separation map — AI design, classical simulation, and quantum-refined steps, cleanly attributed.
- Classical-comparator check — each advantage claim tested against a named baseline (FEP / DFT) at a stated problem size.
- Wet-lab / affinity gap list — missing assays, measured affinities, and validation cohorts, named explicitly.
- Patent-ready evidence log — every line carries a source ID and a confidence score.
- Verdict + validation plan — a conservative call and the recommended experiments to change it.
- Reviewer workflow gate — the package stays in draft until an internal reviewer approves it; the software enforces this.
Sample artifacts and any figures shown in a pilot are Illustrative and use obviously fictional targets and molecules. Numbers in examples do not reflect a real program.
05 · Who it's for
The people who must defend the claim.
Discovery & computational chemistry
Teams evaluating a quantum-chemistry vendor or an internal program who need an honest read on what the quantum step actually contributes — and what the classical baseline already delivers.
IP & legal
Counsel and patent teams who need a defensible evidence trail behind filings — so claim language is backed by source IDs, not optimism.
Investors diligencing a quantum-bio thesis
Funds pressure-testing a quantum-for-discovery claim before committing, with overclaims separated from substantiated, baseline-backed results.
Boards approving R&D spend
Directors signing off on discovery budgets who need a conservative, gap-flagged package rather than a hype-driven roadmap.
Material for investors and boards is decision-support and general information only — not investment, legal, tax, or securities advice, and not an offer or solicitation.
06 · Worked example
What a needs-baseline verdict looks like.
A fictional case, to show the discipline in action.
Illustrative Helixene Bio (fictional) claims "quantum advantage in binding-affinity prediction" for a fictional target, KRAS-X7. Verdict: Monitor
The package shows an in-silico affinity uplift but provides no FEP or DFT comparator, no stated problem size, and no measured wet-lab affinity. Veriqa attributes the reported uplift to an AI/ML scoring step rather than the quantum refinement, records the advantage claim as needs-baseline, and caps maturity. The validation plan is explicit: run a FEP/DFT baseline at a named problem size and an affinity assay before any advantage claim is recorded.Illustrative example with a fictional company, target, and result. Not a real assessment and not indicative of any actual program. Decision-support only — verdicts assist, but do not replace, your team's independent scientific and commercial review.
07 · Questions
Straight answers.
Does Veriqa run quantum chemistry itself?
No. This is a Path A engagement: the Molecular Design Agent grades the evidence behind a design or claim and structures the decision. It does not design or run quantum circuits, integrate quantum hardware, or itself claim quantum advantage. It separates the AI, classical, and quantum-refined steps and tells you what each one actually supports.
Is this generally available?
No. The Molecular Design Agent is at Enterprise pilot stage — engaged under scoped enterprise pilots, not general availability. We are deliberate about where it applies and where it does not yet.
Who reviews the output?
Every package passes through an enforced reviewer workflow gate: it stays in draft until an internal reviewer approves it, and the software enforces that gate. Independent expert review — for example by a computational chemist — is recommended for high-stakes claims; we do not claim that such independent experts currently staff sign-off.
How do you protect our IP and confidentiality?
Pilot inputs are treated as confidential and handled under engagement terms. We do not publish your targets, molecules, or results, and the conservative evidence log is built to support — not jeopardize — your filings. See our Privacy Policy for data handling.
What inputs do you need?
A design dossier or claim package: the candidate molecules or designs, any simulation outputs (FEP, DFT, MD, docking, ADMET, or DFT/ML surrogates), the specific quantum-advantage claim and how it was produced, any wet-lab or affinity data, and the decision you are trying to make. More complete inputs yield a tighter gap list.
Will the pilot expand to autonomous quantum execution?
That is future optionality, not a commitment. Any forward-looking capability or timeline is subject to material uncertainty and may change. Today's engagement is firmly Path A: evidence grading and decision structuring.
Pressure-test a quantum-discovery claim or program.
Bring us a vendor pitch, an internal result, or a draft filing. We separate the methods, demand the baselines, flag the wet-lab gaps, and hand you a conservative, defensible package.
Forward-looking statements about future products, capabilities, timelines, or outcomes are subject to risks and uncertainties and may differ materially from actual results. Veriqa outputs are decision-support requiring your own independent verification and professional advisors; nothing here is investment, legal, tax, medical, or securities advice.