Time since we first saw it
Day 62
Up longer than most. Similar jobs usually come down between day 7 and day 60.
From 12,338 similar postings that came down. Coming down is not always a hire.
Warsaw · Warsaw, Masovian Voivodeship, Poland
Live, first seen by us 62 days agoLocation type: On-siteFrom the employer’s posting
Employment type: Full-timeFrom the employer’s posting
Department: Product & EngineeringFrom the employer’s posting
Build and improve LTV and LTGP models from retention, expansion, usage and cost to serve
Take the LTGP model live as the value signal in Meta and Google bidding
Backtest and calibrate predictions on historical cohorts without leaking future data
Turn predicted LTV and payback into acquisition, segment and pricing recommendations
Measure incremental impact of marketing and sales activity using experiments and causal methods
Build and improve customer LTV and LTGP models.
Take the shadow model to a live value signal in Meta and Google, watch how the platforms respond, and keep improving it.
Backtest on historical cohorts without leaking future information, check calibration and quantify uncertainty.
Turn predicted LTV, cost to serve and payback into recommendations on acquisition spend, customer segments and pricing.
measure the incremental impact of marketing and sales activity. Use experiments and causal methods to guide budget allocation.
Experience with customer lifetime value, survival or retention modeling, and SaaS or usage-based economics.
Our mission is to give every company an AI employee.
In the employer’s words · Read in context · Company website ↗
These dates show when we found the posting. Check the employer’s website to confirm it is still accepting applications.
Day 62
Up longer than most. Similar jobs usually come down between day 7 and day 60.
From 12,338 similar postings that came down. Coming down is not always a hire.
2 open jobs
Open jobs by the week we first saw them, last 12 weeks.
Hiring recordAdd your resume and we’ll mark which of these tools you already use, and point out anything in the posting that could rule you out.
- Connect customer economics to financial planning. Use the customer models to inform growth, profit, runway and fundraising scenarios, with clear sensitivities rather than false precision. - Make the models usable. Build tested, reproducible Python code and customer-lifecycle dashboards that explain the predictions, assumptions and uncertainty, so the team understands the whole customer lifecycle. - Work closely with growth, sales, finance and the Data Engineer. Own the modeling and interpretation; partner on reliable datasets and shared metric definitions rather than owning warehouse infrastructure.
- Depth in experimentation and causal inference. You can distinguish correlation, attribution and incremental impact, and explain what the available data cannot establish. - Advanced SQL and strong Python engineering skills. You can build, test and maintain reproducible modeling code, not just explore data in a notebook. - Commercial judgment. You can turn a prediction into a defensible acquisition, pricing or sales-investment decision, including a recommendation not to act when evidence is weak.
Tech ClickHouse, Hex, Python, PostHog, Stripe data. How We Work
Complete your application on jobs.ashbyhq.com. The employer’s form will show what is required.
Already applied? Track this application
Viktor is the AI employee that lives in Slack and Microsoft Teams, connects to thousands of tools, and does the work: reports, dashboards, code, campaigns. Our mission is to give every company an AI employee.
The team is small. The scope is not.
You predict what customers will be worth over their lifetime, and use those predictions to guide acquisition, pricing and sales investment. This is a quantitative modeling role: you connect retention, expansion and product usage to customer lifetime value, test whether the predictions hold up, and make clear when they are too uncertain to act on.
The first place those predictions land is bidding: we are moving to lifetime gross profit, LTGP, as the value Meta and Google optimize against. A first model runs in shadow today. You take it live and own it.
You'll work onsite in Warsaw alongside Fryderyk, Matt, the growth team and finance, reporting to Fryderyk initially.
Our customers haven't finished their lifetimes. Many cohorts are young, histories are incomplete, and some segments have little data. Meanwhile, new product surfaces change how customers use Viktor, expand and retain. A model that fits yesterday's customers may not predict tomorrow's.
You need to distinguish durable signals from noise, estimate value before a cohort matures, and show how much confidence a decision deserves. Backtesting, calibration and uncertainty are part of the work, not checks added at the end.
Attribution is another hard problem: a channel can appear to bring valuable customers without causing additional value. You'll separate attributed revenue from incremental impact, including the effect of sales activity, so we know where another dollar of investment is likely to pay back.
Build and improve customer LTV and LTGP models. Predict lifetime value and lifetime gross profit from retention, expansion, usage, cost to serve and customer characteristics. Account for incomplete histories, young cohorts and sparse segments; revisit assumptions as customer behavior changes.
Ship LTGP as the live bidding signal. Take the shadow model to a live value signal in Meta and Google, watch how the platforms respond, and keep improving it. By day 90 this is live, not a write-up.
Validate predictions before they drive decisions. Backtest on historical cohorts without leaking future information, check calibration and quantify uncertainty. Monitor model performance and define when evidence is too weak to act on.
Guide acquisition and pricing. Turn predicted LTV, cost to serve and payback into recommendations on acquisition spend, customer segments and pricing. Make assumptions and trade-offs explicit.
Model attribution and incrementality. Connect acquisition spend to customer lifetime value, quantify which channels bring valuable customers, and measure the incremental impact of marketing and sales activity. Use experiments and causal methods to guide budget allocation.
Connect customer economics to financial planning. Use the customer models to inform growth, profit, runway and fundraising scenarios, with clear sensitivities rather than false precision.
Make the models usable. Build tested, reproducible Python code and customer-lifecycle dashboards that explain the predictions, assumptions and uncertainty, so the team understands the whole customer lifecycle.
Work closely with growth, sales, finance and the Data Engineer. Own the modeling and interpretation; partner on reliable datasets and shared metric definitions rather than owning warehouse infrastructure.
Strong statistical and quantitative modeling skills. You can reason about retention, expansion and lifetime value, choose appropriate methods, and explain their assumptions and limits.
You know how to validate a model: backtesting, calibration, uncertainty estimation, leakage prevention and performance under changing customer behavior.
Depth in experimentation and causal inference. You can distinguish correlation, attribution and incremental impact, and explain what the available data cannot establish.
Advanced SQL and strong Python engineering skills. You can build, test and maintain reproducible modeling code, not just explore data in a notebook.
Commercial judgment. You can turn a prediction into a defensible acquisition, pricing or sales-investment decision, including a recommendation not to act when evidence is weak.
Agentic engineering is part of your daily workflow. You use coding agents to move faster while taking responsibility for the methods, code and conclusions.
You work directly with decision-makers, challenge assumptions and follow through on whether your recommendations worked.
Based in Warsaw or willing to relocate, working onsite with the team.
You report to Fryderyk, our CEO, initially. Your mandate spans growth, sales and finance, rather than sitting inside one channel team.
Customer LTV is the primary focus. Your models directly inform where we acquire customers, how we price and where we invest in sales.
Spend decisions sit with Matt, who owns paid acquisition. You sync with him weekly, and your signal is what Meta and Google see.
You work alongside the people making those decisions in Warsaw and see how the predictions perform in practice.
Your models and metric definitions become part of a data layer used by people and AI.
A background in quantitative research, statistics, econometrics or another field where predictions are rigorously tested against outcomes.
Experience with customer lifetime value, survival or retention modeling, and SaaS or usage-based economics.
You've run incrementality tests or built marketing-mix models on real ad spend.
You've built something with LLMs that went beyond a demo.
ClickHouse, Hex, Python, PostHog, Stripe data.
Small team, high trust, low process. Decisions are made by owners, not committees. You will ship your first week. You will talk to users your first day.
We're building an AI employee that does real work across a company's tools.
This is a rare window: everyone here owns something real. Not a task. A surface of the company that customers depend on.
That doesn't last forever. Right now, it's still true.
Top-of-the-market salary and the kind of ownership that only exists at this stage.
This is an onsite role in Warsaw, working alongside the growth team and finance.
New Data Scientist postings at the employers we follow are up 3% over the last 6 weeks compared with the 6 before.
Each bar is one week of new postings. The darkest bars are the last 6 weeks.
Counted from the hiring pages we’ve followed for at least five weeks, so employers we add don’t look like growth. Updated weekly.
We file this job under Data Scientist, by the work the posting describes.
Data scientists use statistics and models to answer harder questions, such as whether a change worked, what demand will look like next quarter, or which customers are likely to leave.
Postings usually ask for Python or R, SQL, statistics, experiment design and common machine learning methods.
Unlike an ML engineer, a data scientist's main output is often an analysis or a first model, not software running inside a product.
Source: US Bureau of Labor Statistics, Occupational Employment and Wage Statistics (May 2025) and Employment Projections (2025 to 2035). Occupation code 15-2051.
Selected passages from the saved posting. Check the full description for conditions and exceptions.
No pay amount identified in the saved description.
Warsaw, Masovian Voivodeship, Poland
Working pattern and location restrictions need checking in the full posting.
No clear work-authorization passage found. Eligibility is unconfirmed.