Data Scientist
Data Scientist at Cummins in Pune builds statistical, machine-learning, and GenAI solutions to business problems: cleaning and profiling data, developing predictive models, and deploying them to production. Core stack is Python, SQL, PyTorch/TensorFlow, Spark, and cloud ML platforms (Azure, AWS, GCP); 5+ years of experience required.
- Compensation
- Not disclosed
- Location
- Pune, Maharashtra, India
- Work setup
- Onsite
- Seniority
- Intern
- Employment
- Full time
- Experience
- 5+ years
- Category
- Data science
- Posted
Skills
Must have
- College/university degree in a relevant technical discipline or equivalent experience
- 5+ years hands-on experience in Data Science, Machine Learning, AI, or related domains
- Advanced proficiency in Python (NumPy, Pandas, scikit-learn)
- Advanced proficiency in SQL
- Strong expertise in supervised/unsupervised learning, predictive modeling, deep learning, NLP, and GenAI
- Experience with PyTorch, TensorFlow, and/or Keras
- Experience with Hugging Face, OpenAI APIs, LangChain, and prompt engineering
- Strong understanding of probability, hypothesis testing, regression, classification, forecasting, time-series analysis, and anomaly detection
- Experience with Spark/PySpark and large-scale data processing
- Experience with Azure ML, AWS SageMaker, and/or GCP Vertex AI
- Experience deploying models using MLflow, Docker, FastAPI or similar
- Knowledge of version control, testing frameworks, CI/CD, and secure development practices
- Proven experience building and deploying end-to-end analytical/ML solutions in production
- Experience processing and managing large, complex datasets
- Experience in Agile software development environments
- Experience validating and testing machine learning systems for reliability and scalability
- Familiarity with business systems, industry requirements, and data governance regulations
- Strong analytical thinking and problem-solving skills
- Effective communication skills presenting technical findings to diverse audiences
Nice to have
- Experience with MLOps orchestration tools such as Airflow or Kubeflow
- Experience with modern data platforms such as Databricks or Snowflake
- Knowledge of vector databases such as FAISS or Pinecone for GenAI applications
- Exposure to RAG architectures and enterprise AI deployment patterns
- Experience with open-source and third-party big data toolsets
- Clustered compute cloud-based implementation experience
Full description
Job Summary:
Solves analytical problems using quantitative approaches through a combination of analytical, mathematical and technical skills. Researches, designs, implements and validates algorithms to analyze diverse sources of data to achieve project specific outcomes by leveraging statistical and predictive modeling concepts.
Key Responsibilities:
Leverages data science methodology to solve business problems. Creates individual algorithms using statistical methodologies through the use of statistical programming languages and tools. Partners with domain experts to verify model capabilities. Implements statistical techniques to clean, prepare and profile the data prior to deeper analysis. Clearly articulates results, methodologies and learnings to stakeholder and peer group. Continuous development and advancement of the team through knowledge sharing and collaboration.
Competencies:
Communicates effectively - Developing and delivering multi-mode communications that convey a clear understanding of the unique needs of different audiences.
Customer focus - Building strong customer relationships and delivering customer-centric solutions.
Decision quality - Making good and timely decisions that keep the organization moving forward.
Manages complexity - Making sense of complex, high quantity, and sometimes contradictory information to effectively solve problems.
Tech savvy - Anticipating and adopting innovations in business-building digital and technology applications.
Data Mining - Extracts insights from data by identifying relationships and patterns through use of a suite of data exploration and data visualization techniques to understand the underlying structure of the data and enable sound conclusions upon model building.
Predictive Modeling - Develops analytical or machine learning models by using appropriate variable transformations, feature selection strategies, imputation strategies, class rebalancing, resampling strategies and quality control measures to generate predictive insights used in solving business questions.
Programming - Creates, writes and tests computer code, test scripts, and build scripts using algorithmic analysis and design, industry standards and tools, version control, and build and test automation to meet business, technical, security, governance and compliance requirements.
Requirements Analysis - Evaluates relationships and interdependencies between requirements based upon their complexity and value to the business in order to determine feasibility and prioritization.
Statistical Modeling - Develops descriptive and explanatory statistical models, and simulations for regression, classification, outlier detection, anomaly detection, time series forecasting using knowledge of foundational statistics such as null hypotheses significance tests, regression models, generalized linear modeling, time series analysis, rank statistics, probability distribution fitting survival analysis, etc. to validate hypotheses for any given statistical or business question.