Relevant Work Experience : 5+ years
Status – Full Time
Location – Pan India (Remote)
Job Description:
Role Overview:
We are looking for a Data Scientist with strong Python-based data science skills who can analyze, evaluate, and improve existing forecasting models and processes. This role is ideal for a hands-on practitioner who enjoys working with time series data, assessing model performance, and translating analytical results into actionable business insights. The focus of this role is not ongoing production support, but rather model accuracy evaluation, data exploration, process enhancement, and stakeholder-facing insights. Prior deep expertise in advanced time-series modelling is desirable but not mandatory; a strong analytical mindset and ability to understand and improve what already exists is key.
Key Responsibilities:
- Forecast Accuracy Analysis & Reporting:
- Calculate, track, and improve forecast accuracy using metrics such as:
- MAPE, weighted MAPE, MAE, RMSE
- Compare forecasts against actuals across multiple time horizons (monthly, quarterly, yearly).
- Create automated and repeatable accuracy reports using Python scripts.
- Visualize forecast performance using box plots, trend charts, and comparative dashboards.
Data Exploration & Model Assessment:
- Perform exploratory data analysis (EDA) on large datasets to identify patterns, anomalies, and data quality issues.
- Investigate and validate data sources used in forecasting (e.g., exchange rates, demand drivers).
- Identify inconsistencies, gaps, or limitations in existing datasets and models.
- Evaluate why current forecasting models may not be widely adopted (e.g., accuracy, interpretability, business mismatch).
Forecasting Models & Techniques:
- Work with and evaluate classical time series approaches such as:
- Naive Forecasts
- Moving Averages
- ARIMA / related statistical models
- Adapt or enhance existing models to incorporate new data channels and business inputs.
- Compare model performance objectively using standardized accuracy metrics.
Business Storytelling & Stakeholder Communication:
- Translate statistical outputs and accuracy metrics into clear business narratives.
- Help define what “good forecast accuracy” means in a practical business context.
- Highlight trade-offs between accuracy, stability, and interpretability.
- Communicate findings and recommendations to non-technical stakeholders.
Automation & Engineering Best Practices:
- Build reusable, maintainable Python code for analytics and reporting.
- Leverage existing codebases, understand their design, and improve them where necessary.
- Apply version control and deployment best practices:
- Git
- Docker
- CI/CD pipelines (basic to intermediate level)
Required Skills & Qualifications:
Technical Skills
- Strong Python programming experience, focused on data science use cases.
- Hands-on experience with:
- Pandas, NumPy
- Matplotlib / Seaborn
- Scikit-learn
- Statsmodels
- Experience performing EDA on large datasets.
- Solid understanding of forecast accuracy metrics (MAE, RMSE, MAPE, etc.).
- Ability to visualize model results clearly and effectively.
Domain & Analytical Skills:
Working knowledge of time series forecasting techniques (experience preferred, not mandatory).
Ability to assess model reliability, interpretability, and business relevance.
Strong problem-solving mindset with attention to data and process gaps.
Professional Attributes:
Ability to quickly understand existing analytical solutions and their business context.
Comfortable working independently with minimal handholding.
Strong written and verbal communication skills.
Curious, improvement-oriented mindset with a focus on business impact.
Nice to Have:
Prior experience in forecasting domains such as finance, demand planning, or exchange rate modeling.
Exposure to cloud-based data environments.
Experience in delivering analytics solutions that drive adoption and decision-making