Data Science Lead
Paystack · Data team (Engineering), reporting to the Data Analytics Lead
- Data Scientist
- Nigeria, Kenya, Ghana, Cape Town or UAE
- 5+ years
At a glance
About the company
Paystack is a payments company. It says over 300,000 businesses across Nigeria, South Africa, Ghana and Kenya use its payments gateway, including MTN and Qatar Airways. It is part of The Stack Group, a family of technology brands in payments, banking and consumer products.
Description
You lead advanced analytics in Paystack's data team, with credit risk for Paystack MFB, its microfinance bank, as the main focus. You will:
- Set the direction and priorities for advanced analytics, and choose the tools the team uses to scale its work
- Build and maintain credit risk models (PD, LGD, EAD, default and segmentation) for Paystack MFB
- Forecast revenue, payment volume and margins, with what-if scenarios for leaders
- Back-test and stress-test models, and document them so auditors and regulators can follow them
- Prepare data for modelling, and work with Data Engineering on automated pipelines and data governance
Qualifications & Skills
Must have 6
- A master's degree in statistics, economics, data science or another quantitative field
- 5+ years in predictive modelling or data science, ideally in fintech or lending
- Proven experience building credit risk models (PD, LGD, EAD)
- Strong statistical modelling and time-series forecasting
- Advanced SQL and expert Python (Pandas, NumPy, scikit-learn, statsmodels)
- Experience running model pipelines, and explaining results clearly to non-technical people
Nice to have 3
- Application and behaviour scorecards that shaped real lending decisions
- dbt, Airflow and GitHub
- Using AI tools to make analytics work faster and better
How to stand out for this job
Written by our team from the employer's post. Work through it in order, then apply.
1What they actually want
Our plain-words read of this job. It explains the job, not the employer's own words.
This is a hands-on lead role inside Paystack's data team, and most of its weight sits on credit. Paystack has a microfinance bank, Paystack MFB, and someone has to build the models that say how likely a borrower is to default, how much would be lost if they did, and how large the exposure is. You will also own business forecasting, so finance and leadership can see where revenue, payment volume and margins are heading under different scenarios. Because these models feed real lending decisions, they must be tested, monitored and written up clearly enough for auditors and regulators.
Paystack wants a senior modeller who can also set direction: five or more years of predictive work, ideally in fintech or lending, a master's degree in a quantitative field, and real PD, LGD and EAD models behind you. Strong Python and SQL are taken for granted. What will set you apart is proof that your models changed lending decisions, and the ability to explain a forecast to a non-technical leader in a few sentences. Scorecard work, dbt and Airflow are useful extras.
2Words to use in your CV
Use the words that are true for you. The dark ones matter most for this job.
- credit risk modelling
- PD, LGD, EAD
- predictive modelling
- time-series forecasting
- Python
- SQL
- scikit-learn
- statsmodels
- back-testing
- stress testing
- model monitoring
- scorecards
- data pipelines
- stakeholder management
- fintech
- lending
4Tailor your portfolio
Make a copy of your portfolio for this job and lead with these three projects. Give the problem, the data, the tools and the result as a number.
- 01
A probability-of-default model
A PD model on a public loan data set, with the features, the validation results and the cut-off you would use.
- 02
A revenue forecast with scenarios
A time-series forecast of monthly payment volume, with base, best and worst cases, and its error on held-out months.
- 03
A model monitoring report
A one-page back-test and drift check for a credit model, written so an auditor could follow it.
3Tailor your CV
- Headline“Lead Data Scientist | Credit Risk Modelling, Forecasting, Python”
- Lead with credit modelsPut your PD, LGD and EAD models first, with the size of the loan book and the decision each one informed.
- Show the model's whole lifeMention back-testing, stress tests, drift monitoring and the documentation you wrote for audit.
- Prove your forecastingGive one revenue or volume forecast and how close it came to the actual numbers.
- Show leadershipName the analytics roadmap or standards you set, and the analysts you mentored.
BeforeBuilt credit risk models for the lending team.
AfterBuilt PD and LGD models for a ₦15 billion SME loan book, cutting 90-day defaults by 18% within two quarters.
Example only. Use your own real numbers.
5Learn the missing skill
SQL and Python for Data Analysis
Live cohort · Starts 4 Nov 2026
₦200,000 or 4 × ₦50,000
6Before you apply
Apply on Paystack's careers page
You'll leave Exalda. Your application goes straight to the employer, and we never charge you to apply.
More data jobs
See all 16 jobsData Scientist, Risk
FlutterwaveA fraud-fighting data science job: build machine learning models that spot fraud and risky payments, then help put them live and keep them accurate.
- Data Scientist
- 3–5 yrs
Data Scientist
ALX Africa · RemoteReally a reporting job: pull data together with SQL and Python, clean it, and keep dashboards teams use every week.
- Data Scientist
- 3–5 yrs
Market Intelligence & Insights Analyst
Renmoney · On-siteA research job for curious beginners: run customer surveys and interviews, watch competitors, and turn what you learn into clear insights.
- Business Analyst
- No research experience needed