Blessing Softtech

Job Title

Data Scientist

Experience

3–5 years applying machine learning and statistical analysis to real problems

Employment Type

Full-time

Location

Pune, Maharashtra, India

Stack

Python, SQL, and modern ML libraries; cloud ML services a plus

About the Role

Blessing Softtech is looking for a Data Scientist to turn data into decisions and working products. You will own problems end to end — framing the business question, sourcing and preparing the data, building and validating models, and getting them into production where they actually affect outcomes.

This is a hands-on role in a small team, so it spans more than modelling: expect to write production-quality code, work directly with stakeholders, and be judged on whether the model changed something in the business, not on offline accuracy alone.

Key Responsibilities

Problem Framing & Analysis

- Work with business and client stakeholders to translate vague problems into well-defined, measurable data science questions.
- Assess feasibility early — is the data available, is the signal there, and is a model the right solution at all.
- Perform exploratory data analysis to surface patterns, anomalies, data quality issues, and relationships.
- Run statistical analysis and hypothesis testing to support or challenge business assumptions.
- Define the success metric for each project before building, and hold the work to it.

Data Engineering & Preparation

- Source data from databases, APIs, files, and third-party systems, and validate its quality.
- Build reproducible data pipelines for cleaning, transformation, and feature engineering.
- Handle missing data, outliers, imbalanced classes, and leakage correctly and document the choices made.
- Write efficient SQL against large tables and optimise slow queries.

Modelling & Machine Learning

- Build, tune,



and evaluate models for classification, regression, forecasting, clustering, or recommendation as the problem requires.
- Select algorithms and metrics appropriate to the problem, and justify the trade-offs made.
- Validate rigorously — proper train/test splits, cross-validation, and checks against baselines and overfitting.
- Apply NLP, computer vision, time-series, or deep learning techniques where the problem calls for them.
- Work with large language models and retrieval-based approaches where they are the practical solution — including prompt design, evaluation, and cost control.
- Interpret and explain model behaviour using feature importance, SHAP, or similar methods.

Deployment & Monitoring

- Package models as APIs or batch jobs and work with engineering to deploy them to production.
- Write clean, tested, version-controlled code — not just notebooks.
- Monitor deployed models for drift, degradation, and data quality issues, and retrain when needed.
- Maintain experiment tracking and model versioning so results are reproducible.

Communication & Delivery

- Present findings and recommendations clearly to non-technical stakeholders, with the caveats stated honestly.
- Build dashboards and visualisations that make results usable rather than merely presentable.
- Document methodology, assumptions, limitations, and decisions for every project.
- Support pre-sales and client conversations with technical input where required.

Required Qualifications





- 3–5 years of hands-on data science experience with models that reached production or drove real decisions.
- Strong Python skills, including pandas, NumPy, scikit-learn, and at least one deep learning framework (PyTorch or TensorFlow).
- Strong SQL — joins, window functions, aggregations, and query optimisation.
- Solid grounding in statistics and probability: distributions, hypothesis testing, confidence intervals, and regression.
- Practical command of the machine learning workflow — feature engineering, model selection, hyperparameter tuning, and validation.
- Ability to choose and defend the right evaluation metric for a given business problem.
- Experience with data visualisation (matplotlib, seaborn, Plotly) and at least one BI tool (Power BI, Tableau, or Looker).
- Proficiency with Git and collaborative development workflows.
- Explicit communication — able to explain a model's behaviour and limits to someone with no technical background.
- Bachelor's or Master's in Computer Science, Statistics, Mathematics, Engineering, or a related quantitative field.

Preferred / Good to Have

- Experience with cloud ML platforms — AWS SageMaker, Azure ML, or Google Vertex AI.
- MLOps exposure: MLflow, Docker, model registries, and automated retraining pipelines.
- Experience building with large language models — RAG pipelines, fine-tuning, embeddings, or vector databases.
- Big data tooling such as Spark, Databricks, or distributed processing frameworks.
- Domain depth in a relevant area — finance, healthcare, retail, logistics, or manufacturing.
- Experience with A/B testing design and causal inference.
- Time-series forecasting at scale, or recommendation systems in production.

Pay: From ₹18,000.00 per month

Benefits

- Paid sick time

Work Location: In person

📌 Data Scientist (Pune)
🏢 Blessing SoftTech
📍 Pune