Her next opportunity bans AI-written applications, so Flow wrote none of her answers. It checked her fit, mapped the form, and planned each part from her own facts.
Clara is made up. The opportunity, its questions and everything Flow says are real.
Part 1 · Is this for her?
Free, about two minutes. Flow read the official page and her CV.
Is this for me?
CLARA NOVAK Senior Machine Learning Engineer · Amsterdam, Netherlands clara.novak@example.com · +31 6 0000 0000 · Czech citizen SUMMARY Machine learning engineer working on ranking and language technology for seven years. I care most about the people a ranking model quietly decides for: the property owner who never learns why their listing dropped. EXPERIENCE Senior Machine Learning Engineer, Reviews & Search Ranking (Sep 2021 – present; senior since Jan 2024) Online travel marketplace, Amsterdam (about 2,000 engineers; reviews in 20 languages) - Led the review-fairness work, in a team of four engineers and one data scientist (2024-2025). - Found that our review-quality model scored machine-translated reviews lower than reviews written in the guest's own language, in 6 of 20 languages. Small independent hotels, whose guests more often write in another language, were pushed down search results as a result. - Built the test that showed it: 3,000 pairs of the same review, original and machine-translated, labelled by 12 native speakers from our customer-service teams in Lisbon, Istanbul and Kraków. - Proposed and shipped the fix: the model now scores each review in the language it was written in. About 40,000 listings were re-ranked; bookings for independent properties in the six languages rose 7% in the A/B test. - Interviewed 14 small-hotel owners with our partner-support team before and after the change. - Wrote the team's guideline for testing ranking features for language bias, now on the launch checklist of three ranking teams. Machine Learning Engineer, Search (Jul 2019 – Aug 2021) Online grocery start-up, Prague (40 people) - Built product search for Czech and Slovak queries: spelling correction and word-form matching for two languages with rich inflection. Zero-result searches fell from 11% to 4%. - First ML hire; set up the team's offline evaluation and weekly error review. EDUCATION BSc, Open Informatics (Artificial Intelligence), Czech Technical University in Prague, 2019. Graduated with honours. Thesis: "Spelling correction for Czech product-search queries" (grade A). TALKS AND WRITING - "When the translation is the bias: ranking reviews across 20 languages", PyData Amsterdam, 2025. - Internal engineering blog: "What our review model thought of Google Translate" (2024), read company-wide. - No academic publications. OTHER - Volunteer mentor, free evening Python course for career changers, Amsterdam (2023 – present). - Languages: Czech (native), Slovak (fluent), English (C2), Dutch (B1), German (A2). - Skills: Python, PyTorch, Hugging Face Transformers, learning-to-rank (LightGBM), Spark, SQL, A/B testing, annotation design.
They say: “To be eligible for an EV grant, all applicants must be 13 years of age or older.”
You shared: “BSc, Open Informatics (Artificial Intelligence), Czech Technical University in Prague, 2019.”
A degree finished in 2019 and seven years of work put you well above 13.
They say: “Tell us your personal story, and how it relates to what you wish to do.”
You shared: “the property owner who never learns why their listing dropped”
Your summary already names who you care about, and it ties to your small-hotel ranking work.
They say: “tell us what is new or unusual in your vision and understanding”
You shared: “Built the test that showed it: 3,000 pairs of the same review, original and machine-translated, labelled by 12”
Your test method shows original insight into a hidden bias, but the CV does not yet state a new idea beyond your employer's model.
They say: “We want to jumpstart high-reward ideas— moonshots in many cases”
You shared: “Proposed and shipped the fix: the model now scores each review in the language it was written in.”
You ship real fixes at scale, but this fix is inside one company and is not a moonshot of your own.
They say: “advance prosperity, opportunity, liberty, and well-being”
You shared: “Volunteer mentor, free evening Python course for career changers, Amsterdam (2023 – present).”
Your free course for career changers shows a social motive, though it is separate from the idea you would propose.
Lead with the review-fairness finding from 2024-2025: you showed a model quietly pushed small independent hotels down, then fixed it. It gives you a concrete personal story and a real problem to build the idea section around.
You shared: “Found that our review-quality model scored machine-translated reviews lower than reviews written in the guest's own”
Nothing to ask this time: Flow writes none of her answers here, so no draft has a gap to fill.
Part 2 · The plan
Their rule on AI help: “Emergent Ventures no longer accepts AI-written applications. We will test all submissions and reject those using AI for anything other than proofreading.” So Flow writes none of her answers. She writes them; Flow plans and checks.
Flow’s summary
You have real evidence of a problem: a ranking model that pushed small hotels down in 6 of 20 languages, and a fix that raised their bookings. The one thing that matters most is to pick an idea that is yours and goes beyond your employer's model, because the CV only shows work done inside the company. Start by choosing that idea, then write down what the 14 hotel owners told you, before you draft anything. The organiser rejects AI-written text, so every sentence must be yours.
Lines she might write, read the way a selector reads them.
“My fix raised bookings for independent hotels by 7%.”
What you shared backs thisYou shared: “bookings for independent properties in the six languages rose 7% in the A/B test.”
“Machine translation bias is hurting small businesses in every language.”
Not in what you shared“I built a tool that audits any ranking system for language bias.”
A stretchYou shared: “Wrote the team's guideline for testing ranking features for language bias, now on the launch checklist of three”
“The hotel owners told me they felt invisible.”
Not in what you shared“My guideline is the standard across the company.”
A stretchYou shared: “now on the launch checklist of three”
The fact
Interviewed 14 small-hotel owners before and after the change
Leads“Tell us about your proposal”
The form asks for a personal story first, and talking to the people affected shows why you care better than a job title.
The fact
Review model scored translated reviews lower in 6 of 20 languages
Leads“How do you describe your idea in a tweet?”
This is the problem in one clear, checkable finding, so it fits the one-sentence format better than the method or the result.
The fact
Your 3,000-pair test with 12 native speakers
Leads“Breakdown of Expenses”
It gives you a real cost model, paid native-speaker labelling, that a reviewer can check line by line.
Left out, and why:
Personal story that relates to what you want to do
Open with these conversations. Write what the owners said that the A/B number did not show, and why that made you want to work on this.
From her CV: “Interviewed 14 small-hotel owners with our partner-support team before and after the change.”
What is new or unusual in your vision and understanding
Explain the finding as something most people would not expect: a quality model treating translation as low quality. Say what you think others miss.
From her CV: “Found that our review-quality model scored machine-translated reviews lower than reviews written in the guest's own”
The problem you intend to solve
Name who is harmed and how. Keep to the 6 of 20 languages the CV shows.
From her CV: “Small independent hotels, whose guests more often write in another language, were pushed down search results as a result.”
Evidence the idea works
Give this once, with its scale and the fact that it came from an A/B test. Then say what your new idea does beyond it.
From her CV: “About 40,000 listings were re-ranked; bookings for independent properties in the six languages rose 7% in the A/B test.”
Ballpark budget with revenue sources and expenses
Base the expense lines on paid native-speaker labelling. Get the real prices and hours yourself; the CV has none.
From her CV: “3,000 pairs of the same review, original and machine-translated, labelled by 12”
Time commitment, full or part time, and partners or supporters
You hold a full-time job now, so say plainly how much time you would give. Name any supporters only if they have agreed.
From her CV: “Senior Machine Learning Engineer, Reviews & Search Ranking (Sep 2021 – present; senior since Jan 2024)”
For each question: what it judges, the lines from her CV to build on, and questions for her to answer herself.
How do you describe your idea in a tweet?
What this judges: Whether you can state one clear, checkable problem in a sentence
From her CV: “Found that our review-quality model scored machine-translated reviews lower than reviews written in the guest's own”
From her CV: “Small independent hotels, whose guests more often write in another language, were”
From her CV: “bookings for independent properties in the six languages rose 7% in the A/B test.”
Tell us about your proposal (1500 words)
What this judges: Whether your story, idea and understanding are real, new and fundable
From her CV: “Interviewed 14 small-hotel owners with our partner-support team before and after the change.”
From her CV: “the property owner who never learns why their listing dropped.”
From her CV: “Proposed and shipped the fix: the model now scores each review in the language it was written in.”
From her CV: “bookings for independent properties in the six languages rose 7% in the A/B test.”
Breakdown of Expenses
What this judges: Whether you can cost the work realistically from experience
From her CV: “Built the test that showed it: 3,000 pairs of the same review, original and machine-translated, labelled by 12”
From her CV: “native speakers from our customer-service teams in Lisbon, Istanbul and Kraków.”
From her CV: “Led the review-fairness work, in a team of four engineers and one data scientist (2024-2025).”
6 ready from her CV12 need her facts6 for her to do
Affected Region
Needs youChoose the region your idea affects. Only you can decide: where the small businesses are, or where you live.
Dedicated Funding Areas
Needs youOptional. Leave blank unless your idea fits a listed tranche; the CV shows none.
First Name
Ready from what you sharedClara
From her CV: “CLARA NOVAK”
Initial
Needs youOptional. The CV shows no middle initial.
Last Name
Ready from what you sharedNovak
From her CV: “CLARA NOVAK”
Suffix
Needs youOptional. The CV shows none.
Email Address
Ready from what you sharedYour email from the CV
From her CV: “clara.novak@example.com”
Email Type
Needs youChoose Personal, School, Work or Other. The CV does not say which this address is.
Phone Number
Ready from what you sharedNumber with country code, as on the CV
From her CV: “+31 6 0000 0000”
Phone Type
Needs youChoose Home, Mobile, Work or Other. The CV does not say.
Twitter (if applicable)
Needs youYour handle, if you have one. The CV lists none.
Country
Ready from what you sharedNetherlands, where you are physically located most of the time
From her CV: “Amsterdam, Netherlands”
State/Province
Needs youSelect your province. The CV names the city but not the province.
State/Province
Needs youSecond State/Province field. Fill it only if it applies to your country; the CV is silent.
City
Ready from what you sharedAmsterdam
From her CV: “Amsterdam, Netherlands”
How do you describe your idea in a tweet?
Do this yourselfOne sentence on your own idea, in your own words. The CV does not name an idea yet.
Tell us about your proposal
Do this yourselfYour story, one consensus view you agree with, your idea, time and partners. All in your own words.
Estimated Budget
Needs youYour own total for the project. The CV has no budget figure.
Breakdown of Expenses
Needs youYour cost lines, such as paid native-speaker labelling. Prices and hours must come from you.
Please upload any other supporting documentation here. (1)
Do this yourselfOptional. No PDFs. Only files you can legally share.
Please upload any other supporting documentation here. (2)
Do this yourselfOptional. Added with 'Add Another Response'.
Please upload any other supporting documentation here. (3)
Do this yourselfOptional. Added with 'Add Another Response'.
Please upload any other supporting documentation here. (4)
Do this yourselfOptional. Added with 'Add Another Response'.
Multimedia URL
Needs youOptional. A link to something public that supports the application. Check your talk is public before you link it.
No idea of your own yet. The CV only shows work inside your employer's model.
Pick one idea that goes beyond it, such as a public language-bias test for ranking systems. Write what is new about it and which small businesses it would help.
before you start writingThe CV does not say what the 14 hotel owners told you.
Go back to your interview notes and pick the two or three things that changed how you saw the problem.
this weekNo budget, revenue sources or time plan.
Price paid native-speaker labelling, add other costs, and decide how much time you would give next to your job.
before you submitUnclear whether you can reuse your employer's data, method or results outside the company.
Check your contract and ask your manager what you may describe publicly. Your PyData talk may show what is already public.
before you submitPart 3 · Before she sends
The fit check is free. The full plan is INR 499 or USD 9, refunded within 7 days if it doesn’t help. A real plan is built from what you share and the real form, and a person reads it once it opens.
Flow’s output as it ran on 28 Oct 2026, unedited. Clara and her CV are made up.