Junior Data Scientist – GenAI & Agentic AI

• *Overall relevant experience:** 5–7 years of overall relevant experience

• *Location:** Pune

• *Work mode:** Work from office

• *Role Summary**

At Emergys, the Senior Data Scientist – GenAI \& Agentic AI independently translates business needs into data science, GenAI and agentic AI solutions. The role blends analytical rigour with rapid prototyping and practical application engineering: framing ambiguous problems, selecting an appropriate solution approach, owning delivery from discovery through validation, and engaging clients and internal stakeholders throughout. The Senior Data Scientist also helps transition successful prototypes into reliable, scalable solutions and mentors junior colleagues.

• *Key Responsibilities**

• Engage with clients and internal stakeholders to understand business goals, frame loosely defined problems, challenge assumptions and translate needs into a solution plan with measurable outcomes.
• Own delivery of analytical and AI workstreams from discovery and data assessment through solution design, prototyping, evaluation, stakeholder demonstration and handover.
• Select the right combination of statistics, predictive modelling, rules, optimisation, retrieval and LLM/agentic techniques based on business context rather than defaulting to the newest or most complex technology.
• Develop robust data pipelines and predictive models using Python and SQL, including feature engineering, model selection, validation, error analysis and interpretation of results.
• Design and implement GenAI applications using LLM APIs, prompt/context engineering, structured outputs, embeddings, RAG and integrations with enterprise data and business logic.
• Build agentic workflows with tool/function calling, API integration, workflow orchestration, state management and controlled multi\-step execution; implement retries, fallbacks, timeouts, guardrails and recovery paths.
• Design evaluation approaches for both predictive models and GenAI/agent systems, including task success, retrieval quality, groundedness, tool selection and execution, reliability, explainability, latency, cost and business value.
• Build rapid POCs and prototypes, test hypotheses with users, communicate findings and trade\-offs to technical and business stakeholders, and recommend whether to iterate, scale or stop.
• Own technical quality for assigned solutions, including modular design, testing, version control, documentation, data handling and readiness for deployment.
• Partner with engineering, product and client teams to move validated prototypes towards production, considering observability, security, access controls, human approval steps, maintainability and operational support.
• Develop reusable components, reference patterns, evaluation datasets and accelerators that improve delivery speed and consistency across engagements.
• Mentor junior colleagues through design reviews, code reviews, modelling guidance and feedback; share practical standards for experimentation and delivery.
• Delivers solutions that address a defined business problem and demonstrates value through agreed operational, financial, customer or productivity measures.
• Owns clear evaluation evidence for predictive quality and GenAI/agent behaviour, including failure analysis and repeatable tests.
• Produces prototypes quickly, incorporates stakeholder feedback and makes evidence\-based recommendations on scaling or discontinuing an approach.
• Transitions validated solutions with appropriate reliability, security, observability, latency and cost controls.
• Improves team capability through mentoring and reusable components, patterns or accelerators.

• *Required Skills and Experience**

• Strong hands\-on Python and SQL skills, with experience building maintainable data science and AI applications.
• Strong foundations in statistics, experimental design, predictive modelling, feature engineering, model selection and validation; experience with regression, classification and clustering as appropriate.
• Experience translating business requirements into analytical or AI solutions and independently delivering work from problem framing through validation.
• Hands\-on experience building LLM applications using APIs and patterns such as prompt/context engineering, structured outputs, embeddings and RAG.
• Practical Agentic AI experience with tool/function calling, API integration, workflow orchestration and state management.
• Experience implementing and evaluating failure handling, including retries, fallbacks, timeouts, guardrails and escalation paths.
• Ability to define fit\-for\-purpose evaluation covering model quality and business value as well as reliability, explainability, latency and cost.
• Experience rapidly developing POCs, validating hypotheses with stakeholders and communicating results and limitations to technical and business audiences.
• Working knowledge of software engineering practices including modular code, testing, Git, API design and documentation.
• Ability to collaborate across client, business, data science, product and engineering teams, and to coach junior colleagues.

• *Preferred Skills**

• Experience delivering GenAI or agentic AI capabilities into production, including monitoring, observability, security controls and human\-in\-the\-loop approval.
• Experience with one or more orchestration frameworks such as LangGraph, LangChain, Semantic Kernel or equivalent; equivalent tools are acceptable.
• Familiarity with evaluation tools such as LangSmith, RAGAS, DeepEval, Arize Phoenix, Promptfoo or equivalent.
• Experience with scikit\-learn and, where relevant, PyTorch or TensorFlow; familiarity with vector databases and retrieval technologies.
• Exposure to cloud platforms, Docker, CI/CD, APIs such as FastAPI, and MLOps/LLMOps practices.
• Experience with cost optimisation, model/provider abstraction, prompt injection mitigation, privacy and access\-control design.
• Domain experience in enterprise operations, manufacturing, supply chain, finance, customer service or other complex business environments.

• *Education**

• Bachelor’s or Master’s degree in Computer Science, Data Science, Industrial Engineering, Operations Research, Economics, or a related field is preferred.