Senior Data Scientist – Credit Risk & Machine Learning

Job Description

Title: Senior Data Scientist – Credit Risk & Machine Learning

Company Name: Sahaj Mobile BD Ltd

Vacancy: --

Age: At least 18 years

Job Location: Anywhere in Bangladesh

Salary: Negotiable

Experience:

• 4 to 8 years
• The applicants should have experience in the following business area(s): Micro-Credit, Financial Technology (Fintech) Startup, Mobile Industry

• Bachelor/Honors

• Background in applied math/statistics/physics or engineering.

• 4 to 8 years
• The applicants should have experience in the following business area(s): Micro-Credit, Financial Technology (Fintech) Startup, Mobile Industry

• Age At least 18 years

Highly Preferred

• Experience in credit risk, lending, or fintech.

• Experience with probability of default (PD), loss modeling, or scorecards.

• Experience using alternative / non-traditional data.

• Strong foundation in statistics, probability, and machine learning·      

• Experience building production-grade models, not just notebooks.

• Experience working with noisy, real-world data.

Ability to independently:

• Define problems

• Build models

• Validate results

• Deploy solutions

SahajMobile is a fintech platform providing smartphone-based consumer financing using alternative data and device-backed collateral. We are building a scalable credit engine for underserved markets and need to significantly deepen our statistical and machine learning capabilities. Currently We are looking for a Senior Data Scientist to design and build credit risk models, statistical frameworks, and data-driven decision systems from the ground up. This role is focused on serious modeling and signal extraction, not just reporting. You will work with large, messy datasets and identify patterns that directly impact underwriting, pricing, and portfolio performance.

Key responsibilities:

Statistical Modeling & Machine Learning

• Build credit risk models (logistic regression, gradient boosting, random forest, etc.)

• Develop default prediction models and probability of default (PD) frameworks.

• Perform feature engineering on structured and unstructured datasets.

• Apply statistical techniques (hypothesis testing, survival analysis, time-series modeling).

  Alternative Data Signal Extraction

• Analyze behavioral, transactional, and device-level data.

• Identify predictive signals tied to repayment behavior.

• Evaluate signal stability and out-of-sample performance

  Model Deployment & Engineering

• Work closely with engineering to deploy models into production.

• Build pipelines for scoring, monitoring, and retraining models.

• Ensure models are scalable and production-ready
  
  Data Infrastructure

• Design clean data schemas and pipelines for modeling workflows.

• Work with large datasets using distributed systems when needed
  
  Model Monitoring & Optimization

• Track model performance over time (drift, accuracy, stability).

• Continuously improve models based on new data and feedback loops

  Tech Stack / Tools: We expect strong experience in most of the following:

• Programming: Python (required), SQL.

• ML / Stats Libraries: scikit-learn, XGBoost, LightGBM, PyTorch or TensorFlow.

• Data Handling: Pandas, NumPy.

• Experimentation & Stats: SciPy, Statsmodels.

• Data Engineering: Airflow, Spark (preferred but not required).

• Cloud: AWS / GCP / Azure.

• Visualization: Tableau, Looker, or similar.
  
  
  

• Festival Bonus: 2

• Mobile Allowance

• Lunch: Fully subsidized