Founder X-Ray report
Alexandra Reyes
Co-founder & CEO · Meridian Robotics · Rotterdam, NL
Alexandra Reyes does not exist. The subject, the company and every number in this report are invented to show the format. Nothing here was measured about anybody.
The engineering record is real and well-received; almost none of it is attached to her name in a place an index or an assistant can read. The company outranks the founder on 34 of 40 queries.
- Generated
- 12 March 2026
- Runtime
- 3m 51s
- Sources read
- 214
- Sections
- 11
Personal Brand Score
This is a demonstration of the LEMANS LABS report format on an invented subject. Alexandra Reyes and the company named here are fictional, every measurement was written for illustration rather than collected, and any resemblance to a real person or company is unintended. Your own report analyses your own public record.
Executive summary
Meridian Robotics is easier to find than the person who runs it. Across 40 neutral professional queries the company page ranks above Alexandra Reyes on 34, and on 12 of them she does not appear in the top ten at all. That is the ordinary shape for a founder of a forty-person company, and it is worth naming rather than assuming: buyers of industrial automation systems research the person before a six-figure commitment, and there is currently no page for them to land on.
The credibility that exists is real, narrow and unlinked. Two conference talks, a seat on a gripper-safety standards working group, and nine direct quotes in trade coverage put her ahead of most founders in the peer set on substance. None of it is reachable from anything she controls: her LinkedIn profile mentions neither the talks nor the committee, and no personal domain exists. The authority signals are sitting in other people's archives.
The AI gap is the largest single number in this report and the cheapest to move. All four assistants fail to name her; one attributes Meridian Robotics to a founder who does not work there. Assistants assemble a person from durable, citable pages, and there are currently three – a conference speaker bio, a university alumni note and a 2023 funding announcement. Two of the three are out of date.
Digital presence
One owned surface, and it is on a platform rather than a domain. The rest of the footprint is company pages, conference programmes and trade coverage – findable, accurate, and written by other people. The footprint is not damaged; it is thin and borrowed.
The name collides with a novelist
An unqualified search for the name returns a Canadian novelist of the same name for the first six results in three of the four indexes we tested. Adding “robotics” or “Meridian” resolves it immediately, which means the collision costs nothing to anyone who already knows where she works and everything to a buyer who was given only a name in a meeting.
Evidence · Search · unqualified name query across four indexes, top-10 results, March 2026
LinkedIn is the whole of the owned footprint
The profile is current, complete and the only page in the top 100 she publishes. It carries the role and the company and omits both conference talks, the standards seat and the technical argument the trade press quotes her for. There is no personal domain, no maintained speaker page and no author page on the company site.
Evidence · Search · top-100 results, 1 self-controlled property; profile audited against the coverage corpus
Two of the three durable pages are wrong
The pages that will still read the same in a year are a 2024 conference speaker bio, a university alumni note and a 2023 funding announcement. The speaker bio gives a job title she held two roles ago and the alumni note names a company that no longer exists. These are the pages assistants and researchers cite first.
Evidence · Search · top-100 results classified by durability; 3 durable, 2 carrying superseded facts
Reputation analysis
Framing is positive everywhere it exists, and it does not exist in much. Of 214 sources, 31 mention her by name and 9 carry a direct quote. We score reputation at 66 rather than higher precisely because the sample is small: a uniformly good record across nine quotes is a weaker finding than a good record across ninety, and we would rather say so than print the average.
The record is good and too small to be a defence
Nine direct quotes across a year is enough to establish a consistent voice and not enough to withstand one bad story. There is no evergreen body of work that a reader lands on if the first thing they find is a complaint, a disgruntled integrator or a competitor's comparison page.
Evidence · News corpus · 214 sources, 31 naming her, 9 with direct quotation, Mar 2025 – Mar 2026
Online visibility
She is findable if you already know the company, and largely invisible if you do not. Six of forty professional queries return her above Meridian Robotics; twelve do not return her at all. The pattern is consistent: queries naming the company find her, queries describing the work find the company.
- The six she wins are all interviews
- Every query where the person outranks the company resolves to a podcast episode or a recorded interview. These are the only surfaces where she speaks in her own words for long enough to rank, and all six sit on somebody else's domain.
- The twelve she loses are the buying questions
- “Who is solving mixed-case palletising”, “bin-picking accuracy benchmarks” and ten siblings return integrators, analysts and two competitors. These are the queries a buyer runs before they know any vendor's name, and the founder is absent from all of them.
- The footprint does not move
- Result volume for the name is flat across the year, with two spikes around the conference talks that decay inside three weeks. There is no ongoing publication, so there is nothing generating new surface area between events.
AI search presence
This is the weakest dimension in the report by seventeen points and the most tractable. No assistant of the four names her unprompted, one misattributes her company to another person, and the two that describe Meridian Robotics at all build the description from the 2023 funding announcement.
One assistant names the wrong founder
Asked who founded Meridian Robotics, one assistant returns a name that belongs to a different automation company entirely and states it without qualification. This is the single most damaging finding in the report: a confident wrong answer is worse than an absence, because a reader has no reason to check it.
Evidence · Assistant queries · direct founder-attribution prompt across four assistants, March 2026
Two assistants describe a company two years old
Both build their answer from the 2023 funding announcement, which is the most-cited durable page about the company. They report the headcount, the product line and the funding position as they were then. The models are not wrong about their sources; the sources are the newest durable pages that exist.
Evidence · Assistant queries · 12 neutral prompts, answers traced to cited sources, March 2026
No canonical description exists to prefer
There is no page anywhere stating who she is, what she works on and why, in her own words, on a stable address. Assistants prefer a canonical source when one exists and assemble one from fragments when it does not. Here there is nothing to prefer.
Evidence · Assistant queries · four assistants; Search · top-100 results, 0 canonical personal pages
Asked who is solving mixed-case palletising, it names three larger vendors and no individuals. Asked directly about Meridian Robotics, it describes the company from its 2023 funding coverage and says it does not have information about who runs it.
Does not name her on any of the twelve category prompts. Asked directly, it describes Meridian Robotics accurately for 2023, states that its information may be out of date, and declines to name a founder rather than guessing.
Names a founder of a different automation company as the head of Meridian Robotics, without qualification. This is the misattribution recorded in section 05 and the most damaging single answer in the set.
Returns the company website and the 2023 funding announcement as citations and reports that leadership is not stated in its sources. It is the only assistant of the four that shows a reader why it cannot answer.
Professional positioning
There is a clear technical argument – that placement accuracy under variable load, not cycle speed, is what decides whether warehouse automation pays back – and it is consistent everywhere it appears. It appears in nine places, and seven of them are somebody else's article. The position is real and unowned.
- What the record actually says
- Across 214 sources the three phrases most often attached to the name are placement accuracy, variable-load tolerance and payback period. That is a systems-engineering identity rather than a founder-story one, and it is unusually specific for a company of this size.
- The argument is quoted, never published
- Nine sources carry the accuracy-over-speed argument. Two are her own conference talks; seven are trade journalists paraphrasing her. There is no article, post or page under her control that states it, so there is nothing for anyone to cite or link to.
- The company positions on price
- Meridian Robotics' own site leads on total cost and integration time. The founder's argument is more differentiating than the company's and appears nowhere on the company site, which means the strongest positioning asset in the business is not in the sales path.
- No view on the category's open question
- The live argument in warehouse automation is whether general-purpose manipulators displace task-specific cells. She has a defensible position on it, stated in one podcast, and no written record of it. This is the highest-authority piece of writing available to her and it does not exist.
Growth opportunities
Upside that is reachable with the business you already have – ordered by how much of it is within your own control.
A single owned page would rank inside a quarter
Nothing she owns competes for her own name, and 17 outlets are already writing about the work. One maintained page – role, the accuracy argument, the talks, the standards seat, dated – enters the top ten for the qualified name almost immediately, because the competition is a novelist and a stale alumni note.
Evidence · Search · top-100 results, 1 self-controlled property, 17 distinct outlets, March 2026
The misattribution is correctable, and correcting it compounds
One assistant names the wrong founder because the durable pages do not say otherwise. Publishing a canonical page and updating the two stale sources fixes the input all four assistants read, and the AI presence dimension is the lowest of the six by seventeen points, so it also moves the headline most.
Evidence · Assistant queries · 4 assistants; Search · 3 durable pages, 2 carrying superseded facts
The standards seat is a procurement asset going unused
Sitting on a gripper-safety working group answers the safety question a buyer will ask anyway, before they ask it. It is currently documented in a committee roster and nowhere else. Naming it on the company site and the profile costs an afternoon.
Evidence · Search · committee membership traced to 1 source, absent from all owned surfaces
Twelve buying-intent queries have no founder in them
The queries a buyer runs before they know a vendor return integrators and analysts. Writing the accuracy argument as a plain technical piece puts a named engineer into a result set that currently contains only intermediaries.
Evidence · Search · 12 category and problem queries with no top-10 result for the name, March 2026
Risks
Structural exposure visible in the public record. None of these are predictions; each one is something already observable.
A confident wrong answer about who runs the company
One assistant attributes Meridian Robotics to a founder who does not work there and offers no hedge. A buyer or a journalist checking quickly gets a wrong fact stated plainly, and nothing in the current footprint contradicts it.
Evidence · Assistant queries · direct founder-attribution prompt, four assistants, March 2026
Reputation rests on nine quotes
The record is uniformly positive and very small. There is no durable body of owned writing to absorb a bad story, and a single unhappy integrator posting a detailed complaint would rank against a footprint of 31 mentions with no counterweight.
Evidence · News corpus · 214 sources, 31 naming her, 9 quoting her, Mar 2025 – Mar 2026
Platform dependency on one profile
The single owned surface is a profile on a platform. If it changes shape, is restricted or is lost, the owned share of the footprint goes to zero and every fact about her online is one somebody else wrote.
Evidence · Search · top-100 results for the name, 1 self-controlled property, March 2026
The differentiating argument sits outside the sales path
The accuracy-over-speed position is more persuasive than the company's price-and-integration message and appears nowhere a prospect will encounter it. Competitors are free to make the same argument first, and one of them already ranks for the benchmark queries.
Evidence · Search · company site audit against the coverage corpus; 9 sources carrying the argument, 7 third-party
Strategic recommendations
What we would do, in order, if this were our own position to defend.
Publish one canonical page on a domain you control
Role, company, the accuracy argument in three paragraphs, both talks with links, the standards seat, and a dated update line. It becomes the citation target for the 17 outlets already covering the work and the source all four assistants currently have to guess at.
Evidence · Search · top-100 results, 1 self-controlled property; Assistant queries · 0 of 4 naming her
Correct the two stale durable pages first
The conference speaker bio and the alumni note are the two most-cited stable pages and both carry superseded facts. Both are editable through ordinary channels. Fix them before publishing anything new, because they are what the assistants are reading.
Evidence · Search · 3 durable pages in top 100, 2 carrying superseded facts, March 2026
Write the accuracy argument once, properly
Seven journalists have paraphrased it and no version exists in her own words. One technical piece with the load-tolerance numbers behind it serves the twelve buying-intent queries, the category question and the company's own sales path at the same time.
Evidence · News corpus · 9 sources carrying the argument, 7 of them third-party paraphrase
Put the founder in the company's own funnel
The company positions on cost; the founder positions on accuracy, and the second is the one that differentiates. Add a named engineering-view page to the company site and link it from the product pages, so the strongest argument in the business is reachable from the sales path.
Evidence · Search · company site audit; founder positioning absent from all company product pages
Qualify the name everywhere it is published
The unqualified name returns a novelist. Every bio, byline and speaker submission should read “Alexandra Reyes, Meridian Robotics” rather than the name alone, so that the qualifier travels with every new page and the collision stops mattering.
Evidence · Search · unqualified name query across four indexes, novelist in first 6 results in 3 of 4
90-day action plan
Three windows. The first pays for the report, the second compounds it, the third only works once the first two are done.
Correct the four stale role descriptions
Two durable pages and two company-adjacent listings carry a superseded title or employer. All four are editable by request. Start with the conference speaker bio, which is the most-cited page about her.
Add the standards seat and the talks to owned surfaces
Profile and company site, one line each with links. This is the strongest procurement credential in the report and it currently exists only in a committee roster.
Stand up the canonical page
One page on a domain under her control: role, the accuracy argument, both talks, the standards seat, a dated update line. Ship it thin and correct rather than complete.
Publish the accuracy argument as a technical piece
The load-tolerance case, with the numbers, in her own words, on the canonical page and cross-posted to the company site. It is the only asset that serves the buying-intent queries and the category question together.
Re-test assistant attribution and iterate
Re-run the twelve prompts after the canonical page has been live a month. Expect the misattribution to resolve first and the unprompted naming to lag by a quarter; if the wrong founder persists, the stale sources feeding it have not all been found.
Build a second and third durable page
Three durable pages is a footprint that any single bad story outranks. Two more substantive pieces – the category position and a build-versus-buy piece – take the durable slice from three to five and give a reader somewhere to land.
Personal Brand Score
The engineering record is real and well-received; almost none of it is attached to her name in a place an index or an assistant can read. The company outranks the founder on 34 of 40 queries.
Personal Brand Score
47/100
The headline score is the plain mean of the six dimensions below. No weighting, no curve.
The name resolves, but to a warehouse-automation vendor rather than a person – 6 of 40 professional queries return her above the company.
A complete and current LinkedIn profile carries the footprint almost alone; there is no personal domain and no maintained bio anywhere else.
Two conference talks and a standards-committee seat are strong signals, and neither is reachable from any page she controls.
Uniformly positive across 214 sources, and thin: 31 of them mention her by name and 9 quote her, so the number rests on a small sample.
No assistant of four names her unprompted; one names the company and attributes it to a different founder entirely.
The pick-and-place accuracy argument is consistent wherever it appears, and it appears in nine places, seven of which are other people's pages.
Keep reading
Three more, same rubric.
The rubric only means something across subjects. Read a second one and the shape of the first starts to explain itself.
Automotive & energy
Tesla
Category-defining brand, thinning narrative control.
Illustrative demonstration, not a real assessment.
Financial infrastructure
Stripe
Developer trust at ceiling, buyer-side story unfinished.
Illustrative demonstration, not a real assessment.
Artificial intelligence
OpenAI
Maximum awareness, contested positioning.
Illustrative demonstration, not a real assessment.
Now see what yours says.
Same format, same rubric, more sections than this one – applied to your own public record. About six minutes from here.
- One-time payment – no subscription
- Readable for 90 days
- Export to PDF, yours to keep
- Secure processing via Stripe