Data Scientist / Quantitative Risk Analyst
Madfish
- London, England, United Kingdom
- London, England, United Kingdom
About
autonomy with accountability , and maintain a culture where clear documentation beats hierarchy.
What you’ll do
Engineer risk‑focused features
(borrower, lender, property, geography) in Python/PySpark.
Develop and validate
PD / LGD models using WoE, IV, logistic GBM, XGBoost, or similar.
Prototype lender‑health metrics
(capital‑diversification, portfolio turnover, market concentration, etc.) for client dashboards.
Create robust, reproducible
data pipelines
(git‑versioned, unit‑tested, CI in GitLab).
Produce concise notebooks & dashboards that can feed automated PDF reports.
Must‑have qualifications
4 – 6+ years in data science, risk analytics, or credit‑modeling.
Strong Python (pandas, NumPy, scikit‑learn)
and
SQL; solid PySpark on distributed data a big plus.
Hands‑on experience building or validating
credit‑risk or fraud
models (PD, scorecards, Basel/IFRS 9, etc.).
Fluency in statistics (inferential tests, multicollinearity, model monitoring).
Git workflow, code review discipline, and comfort with Agile/Kanban boards.
Clear written & spoken English; able to summarize findings for non‑technical stakeholders.
Nice‑to‑haves
Familiarity with U.S. mortgage or private‑lending data.
Experience with Postgres, MinIO/S3, or dbt.
Knowledge of BI/visualization tools (Plotly, PowerBI, Looker, etc).
Prior work in a fully remote, internationally‑distributed team.
How we work
Stack:
Python • PySpark • PostgreSQL/Snowflake • GitLab CI • AWS & on‑prem Spark
Communication:
Slack, Zoom, Notion. Meetings kept lean; deliverables drive the schedule.
Culture:
Low‑ego, high‑ownership. We favor clarity, rapid feedback loops, and well‑documented processes.
#J-18808-Ljbffr
Languages
- English
Notice for Users
This job comes from a TieTalent partner platform. Click "Apply Now" to submit your application directly on their site.