Example scan · frozen output · free-scan route R0 R3 R5 R7
We planted one impossible claim. First Landed caught it.
This is a real production scan of a synthetic CV for Daniel Lim, frozen as returned by the engine.
Disclosure. Synthetic test case: no real candidate is depicted. We planted the 2008 desk-leadership claim to test whether the engine catches a chronology problem. Every sentence inside the numbered result blocks is the engine’s own output for that CV against one JD. The grey notes around them are ours.
Captured 2026-09-04T06:13:43Z on deepseek-v4-flash, order FL-20260904-aa0c9666. Frozen output: run the same CV again and the wording will differ.
Risky claims
3 lines·2 need proof·1 cut
This is the interview attack surface, what a careful reviewer is most likely to push back on, not a list of mistakes. The candidate should read it as preparation.
HIGH SEVERITY chronology conflict: 2008-2011 BSc Economics at University of Bristol versus "LED THE ENTIRE LONDON FX TRADING DESK DURING THE 2008 FINANCIAL CRISIS, MANAGING A USD 50 BILLION BOOK AND DELIVERING USD 300 MILLION P&L IN THE FIRST QUARTER."
Cut- Why a reviewer pushes back
- The claim requires Daniel to have led a London FX desk during the 2008 financial crisis while the CV shows him as a full-time undergraduate in Bristol from 2008 to 2011. There is no part-time, placement, or internship explanation anywhere in the CV, and no employer is named for this role. The USD 50 billion book and USD 300 million P&L figures are extraordinary for any trader, let alone a student, and no measurement window or audit trail is provided.
- Smallest proof
- A signed employment contract or desk mandate letter from the named employer covering 2008, plus a P&L statement or broker statement for Q1 2008, and a letter from the university confirming the degree was completed part-time or with a placement year.
"Senior trader, StraitsBridge Bank, 2021-2026. Led desk research sessions, reviewed model assumptions with technology, and mentored two analysts."
Needs proof- Why a reviewer pushes back
- "Led desk research sessions" and "reviewed model assumptions" are vague seniority claims that do not establish research ownership. The JD asks for evidence of owning a model lifecycle from hypothesis to desk feedback; this bullet describes participation, not ownership. The 2021-2026 span also extends to the present with no end date, which is fine, but the promotion from trader to senior trader at the same bank is not explained with any performance rationale.
- Smallest proof
- A research session agenda or meeting notes where Daniel proposed a hypothesis, showed test results, and documented trader feedback, plus a promotion letter or performance review that names the specific achievements justifying the senior title.
"FX trader, StraitsBridge Bank, 2015-2021. Priced G10 and Asia FX spot and forwards, maintained risk explain notes, and partnered with quants on backtesting carry and momentum screens."
Needs proof- Why a reviewer pushes back
- "Partnered with quants on backtesting" is a supporting role, not a research role. The JD requires the candidate to test signals and review performance decay; this bullet suggests Daniel was a user of quant output, not a producer. The risk explain notes are a good communication artifact but are not described as research publications.
- Smallest proof
- A copy of one risk explain note that shows Daniel explaining a P&L move with a clear hypothesis and data, plus an email thread or document showing his specific contribution to a carry or momentum backtest.
Result
Fit score
42/100
Verdict
Needs more evidence
Independent same-language re-runs were stable: English 42/42 and Simplified Chinese 28/28. This capture shows real cross-language divergence: English returned 42, “Needs more evidence”; both Chinese runs returned 28, “Skip”.
Summary
The score reflects a genuine partial overlap: Daniel has real Python and SQL skills, direct FX and rates product knowledge, and years of trader-quant partnership experience that map to the JD's communication and collaboration requirements. The single biggest factor keeping the score down is the absence of any evidence that Daniel personally owned a model lifecycle from hypothesis to desk feedback; his CV describes partnering, reviewing, and mentoring, but never leading a research project end to end. Additionally, the CV contains a dated claim that cannot stand as written, which materially weakens overall credibility.
Score breakdown
| Criterion | Score | Max | Why |
|---|---|---|---|
| Hard requirements coverage | 18 | 35 | Strong Python and SQL are claimed but not demonstrated with specific projects; FX and rates product knowledge is solid and directly evidenced by six years of trading G10 and Asia FX. |
| Experience relevance | 20 | 30 | Trading experience is highly relevant to the JD's trader partnership requirement, but the CV shows no quant research role, no signal testing, and no model lifecycle ownership. |
| Evidence quality | 3 | 20 | The CV is almost entirely role descriptions without metrics, outcomes, or artifacts; the one concrete metric (USD 50 billion book, USD 300 million P&L) is attached to a chronology conflict. |
| Claim defensibilityBiggest gap | 1 | 15 | The 2008 London desk leadership claim directly conflicts with the 2008-2011 Bristol undergraduate degree and cannot be defended as written. |
Each bar shows the captured score against that factor's maximum. The tag follows the product's lowest score-to-maximum ratio.
Strongest angle
Daniel's six years as an FX trader at StraitsBridge Bank, pricing G10 and Asia FX spot and forwards while partnering with quants on backtesting carry and momentum screens, is the most defensible match to the JD's requirement for FX product understanding and trader collaboration.
Missing evidence
Any specific Python or SQL project: a data pipeline, a backtest framework, a signal analysis notebook, or a database query set with measurable outcomes.
Evidence of owning a model lifecycle: a named signal or strategy Daniel proposed, tested, refined, and handed to a desk with documented feedback.
A written research sample: a memo, a market note, or a presentation that demonstrates the "clear written communication in English" the JD requires.
Any quant research role title or formal research responsibility; "partnered with quants" is not the same as conducting research.
A plausible explanation for the 2008 London desk leadership claim, including dates, employment, and reporting structure.
Fact-Checker refused the flagged claim
The lines that need work here are the ones flagged above; they need removing or evidence, not rewording.
The 2008 London desk claim must be cut or fully documented, and the "partnered with quants" and "led desk research sessions" bullets need artifacts before they can be sharpened. The graduate analyst bullet from 2013-2015 is the only weak-but-honest line that can be safely rewritten.
Before
"Built VBA and Python scripts for daily P&L explain, yield-curve checks, and FX forward-point sanity checks."
After
"Built Python and VBA scripts for daily P&L explain, yield-curve checks, and FX forward-point sanity checks, supporting the desk's market data analysis workflow. [NEEDS PROOF: specific script or notebook that shows the Python implementation and the data sources used]"
Next step
Do not apply yet; gather a written research sample, a specific Python or SQL project artifact, and a documented example of a model or signal you personally owned from hypothesis to desk feedback, and remove or fully evidence the 2008 London desk claim before submitting.
What you saw
Fair credit, then a hard stop.
The useful signal is not a low score. The scan gave credit for real FX trading experience, then isolated the one impossible 2008 claim as an interview-killing risk. It shows fair scoring, specific evidence gaps, and no permission to keep an indefensible line.
Paste one CV and one JD. The scan is built to surface fit, missing proof, and claims you should not carry into an interview.