17 Nov
|
Snapdeal
|
Gurugram
ABOUT THE ROLE
Solve real, large-scale e-commerce problems with data. You will work with production datasets spanning customer behaviour, catalog,
pricing, and logistics building predictive models and insights that shape business decisions. You will partner with Data Science, Product, Category, and Engineering teams to frame problems, build models, and turn findings into recommendations leaders can act on. We look for an AI-first approach: using GenAI to accelerate analysis and automation, not as an afterthought.
WHAT YOU'LL DO
- Write and optimise complex SQL to extract and aggregate data across large transactional and behavioural datasets.
- Run exploratory analysis to surface patterns, anomalies, and opportunities across the customer and seller lifecycle.
- Build and evaluate predictive models in Python demand forecasting, conversion propensity, churn, RTO prediction, delivery date estimation, or fraud detection.
- Clean, transform, and engineer features from messy real-world data; define metrics and validate models rigorously.
- Use GenAI tools and LLM APIs to accelerate exploration, code generation, and insight summarisation.
- Automate repetitive analytical work through reusable Python scripts, notebooks, and dashboards.
- Present analyses and recommendations to technical and non-technical stakeholders with data-backed reasoning.
WHAT WE'RE LOOKING FOR
- SQL — complex joins, window functions, CTEs, and aggregations over large datasets.
- Python, Pandas & NumPy — clean analytical code; hands-on data wrangling, merging, reshaping, and vectorised operations.
- scikit-learn & predictive modelling — the full workflow — train/validation splits, feature engineering, tuning, and evaluation; regression, classification, and tree-based methods (Random Forest, XGBoost/LightGBM) — plus the judgement to know when a simpler model wins.
- Statistics — distributions, sampling, correlation vs. causation, hypothesis testing.
- AI-first mindset — hands-on with ChatGPT, Claude, Gemini, Cursor, or Copilot — and the judgement to validate AI output rather than accept it at face value.
- Ways of working — structured problem-solving, scepticism toward your own numbers, and the ability to explain a model to someone who doesn't build models.
Good to have: Matplotlib/Seaborn/Plotly or BI tools; Git and Jupyter/Databricks; cloud platforms (AWS, GCP, Azure); PySpark; personal projects, Kaggle, or open-source contributions.
ELIGIBILITY
Pursuing a Bachelor's or Master's in Computer Science, Data Science, Statistics, Mathematics, Economics, Engineering, or a related quantitative discipline. Final-year students and recent graduates preferred, with coursework or projects in machine learning, statistics, or data analytics.
WHAT YOU'LL GAIN
- Production-scale data from one of India's leading e-commerce platforms, serving Tier 2 and Tier 3 India.
- Exposure to the full data science lifecycle — problem framing, modelling, evaluation, and rollout.
- Mentorship from experienced Data Science, Product, and Business leaders, on live projects affecting millions of users.
- Consideration for full time opportunities based on performance.
📌 Data Science Intern (Gurugram)
🏢 Snapdeal
📍 Gurugram