This is what a fit report looks like when it is not flattering you. The founder’s own CV, read once into records, was run against two of the invented postings in the paste screen on 18 September 2026. Nothing below was edited after the runs except that employers, the university and collaborators are described rather than named. Every status is the judge’s own.
The record
A PhD in biomedical engineering, followed by about three and a half years in industry: an instrument company, then a diagnostics startup where the work was immunoassay development under ISO 13485 design control, verification and validation against CLSI guidance, transfer to manufacturing, and a team of up to seven. A year of independent consulting after that. Eighteen publications. No certifications or licences. The record has 5 positions, and the judge cites them by name in every line below.
Posting one: 76, apply
Senior Scientist, Assay Development, at an invented diagnostics company. Twenty-one requirements: nine must-haves, nine responsibilities, three nice-to-haves. Score 76 of 100, verdict strong, recommendation apply.
The nine must-haves
metPhD in biochemistry, molecular biology, analytical chemistry, or a related field, with three or more years of industry experience
A PhD in biomedical engineering, a related field, and about three and a half years in industry at an instrument company and a diagnostics startup, plus a year of consulting.
metDemonstrated experience designing verification studies against CLSI documents such as EP05, EP06, EP07, and EP17
EP05, EP06, EP07 and EP17 named by number on a real product program, under ISO 13485 design controls, with acceptance criteria set by DOE. The strongest single match in the record.
metExperience with antibody pair screening and calibration curve design
Antibody and conjugate selection is explicit at the startup. Calibration curve design is not named as such; the EP06 linearity and EP17 limit work under design control covers the substance of it.
metHands-on immunoassay development: ELISA, chemiluminescence, or lateral flow
An or-list, and the record carries more than one part: ELISA, lateral flow, AlphaLISA and bead-based formats, and a full assay chemistry re-optimised onto lateral flow at matched performance.
metA master's degree with six or more years of industry experience is equally welcome
The alternate path; cleared through the PhD route instead.
metWorking knowledge of ISO 13485 and design control in a regulated diagnostics environment
Two years validating methods under ISO 13485 design control, with stability studies, lot-release testing and CAPA. Noted: the product was a breath-testing device, not an IVD cleared for clinical use.
metExperience with JMP, R, or Python for analysis
JMP evidenced twice in use, for DOE optimisation and for setting V&V acceptance criteria. Python and R are untested skill tags; JMP alone carries the or-list.
metStatistical fluency with method comparison and precision data
EP09 method comparison and EP05 precision both in the V&V design; DOE-driven optimisation drove %RSD below 3%.
partialClear technical writing, as work will be read by regulators
Eighteen publications, eight as first author, SOPs, transfer protocols and validation documentation. The qualifier is not evidenced: the writing stayed inside a design control system with no submission behind it.
Eight met, one partial, no gap among the must-haves, which is why the strong verdict stands: the cap that takes a strong down to moderate fires on a missing must-have, and none is missing. The one partial is worth reading twice. The requirement is not “writes clearly”; it is “writes what regulators read”, and the qualifier is the part the record cannot show. That is the rule on How scoring works: an environment named in a requirement is part of the requirement.
The three nice-to-haves
partialExperience transferring an assay into volume manufacturing
The transfer is real, including a production contamination root cause. No record gives units, lots or rate, so the volume is unknown.
partialExperience with qPCR or isothermal amplification chemistry
qPCR and RT-qPCR as supporting work during the PhD, which ended in 2020. Exposure, not chemistry development; nothing on isothermal methods.
gapExposure to IVDR technical documentation, or to an FDA 510(k) submission
IVDR, 510(k) and 21 CFR 820 appear only as bare skill tags with no accomplishment behind them. Design control at the startup is adjacent, and is not a submission.
Of the nine responsibilities, four came back met (owning V&V design, owning two assays end to end, transfer to manufacturing, mentoring) and five partial, every one of the five for the same reason: the record shows the work but not the design review, the regulatory reviewer or the submission the responsibility names. The summary the judge wrote ends: expect questions on scale and on submission exposure. That is the cover letter’s job.
Posting two: 23, skip
Clinical Operations Manager, Multi-Site Trials, at an invented trials organisation. Twenty-one requirements again: ten must-haves, eight responsibilities, three nice-to-haves. Score 23, verdict poor, recommendation skip.
The ten must-haves
partialDemonstrated budget ownership (forecasting, defending, and being accountable for a study budget), not just budget input.
A $250K-a-quarter budget owned and external testing costs cut by about 45%. Missing: a study budget, and any record of forecasting and defending one.
partialComfortable saying no to a site that is not ready.
Go/no-go recommendations in technical due diligence for investors are direct evidence of making and defending a negative call. Not a site-readiness decision.
partialDirect line management experience, including having had a performance conversation.
Led a team of up to seven and mentored junior researchers. A performance conversation is a specific act and the record never mentions one.
gapConfident with a CTMS.
LIMS and Benchling are listed and analytical platforms have been custom-programmed, so lab software is picked up fast. A CTMS is not in the record.
gapFive or more years in clinical operations.
Eight-plus years since 2016 are in biosensor, assay and analytical development. No position in the record is a clinical operations role.
gapAt least two years running multi-site studies.
Clinical samples analysed across three academic medical centres is multi-site scientific collaboration, not running multi-site studies as the operational owner.
gapWorking knowledge of ICH-GCP.
Fluent in ISO 13485, FDA QSR, 21 CFR 820 and CLSI, which is regulated-environment literacy. ICH-GCP governs trial conduct and appears nowhere.
gapWorking knowledge of the UK Clinical Trials Regulations.
IVDR is listed, which is EU device regulation. The career is US-based and no UK trial framework appears.
gapExperience of an inspection or a sponsor audit.
Work to auditable standards under design control, yes. Being through an inspection or a sponsor audit, no record.
gapA current registration or certification relevant to the role, such as RN, RGN, or an ACRP or SOCRA certification, is required and must be current at point of hire.
The record holds zero certifications. A PhD does not substitute, and this requirement is met only by holding it.
Seven gaps among ten must-haves. Three partials, and each partial is honest about what carries it: real budget ownership, real negative calls, real team leadership, none of it in the setting the posting names. The last line is the one that decides it on its own. A registration the role requires is met by holding it and by nothing else, and the record holds none.
The judge’s summary calls the record a strong assay and diagnostics scientist applying into clinical operations, and says the two do not overlap much. Then: expect a fast no here. The responsibilities agree: two partial, six gap. Nothing in the nice-to-haves changes it; one is met (therapy-area breadth), which is the kind of line a keyword tool would count and a reader would not.
What the two verdicts share
The same record produced both, and the same rules did. In each line the judge names the evidence it used, says what the requirement asked for beyond that evidence, and stops. Where a skill sits in the record as a bare tag with no work behind it (IVDR, 510(k), Python), it says so, in both runs, rather than counting the tag. Where a requirement is an or-list, one part carries it and the line says which.
A skip is the product working, not failing. Posting two would have cost an evening and returned a silence. The verdict cost a few minutes and returned the reason, which is the difference between a search that teaches you something and one that does not. The rule for deciding by hand is on Should I apply?.
Run yours
Paste a posting you are weighing and attach your CV. The first fit is free, no card, and the CV file is deleted after it is read. You get the same table as above, for your record, with the reasoning on every line.