Mastercard Data Scientist I Recruitment 2026: Generative AI Role in Pune

Mastercard is hiring for the position of Data Scientist I in Pune, Maharashtra, with a focus on applying generative artificial intelligence to financial risk and compliance. The role sits within Global Risk and Compliance and centres on developing AI-powered solutions for Anti-Money Laundering (AML) transaction monitoring.

This opportunity is particularly relevant to candidates who can build more than experimental AI prototypes. The job description highlights retrieval-augmented generation (RAG), large language models (LLMs), AI evaluation, model fine-tuning, cloud platforms and the integration of AI applications into production systems.

The position is listed as full-time under Job Requisition ID R-278159. Mastercard has not specified the salary, required years of experience, application deadline or detailed interview process in the supplied listing.

At a glance

• Company: Mastercard
• Position: Data Scientist I
• Job ID: R-278159
• Location: Pune, Maharashtra, India
• Employment type: Full-time
• Team: Global Risk and Compliance
• Primary focus: Generative AI for AML transaction monitoring
• Education: Bachelor’s or master’s degree in Computer Science or equivalent
• Salary: Not specified in the supplied job description
• Experience requirement: Not specified

What makes this role different from a general data science job?

The central objective is to support AML transaction monitoring with AI-based solutions. Anti-Money Laundering programmes help identify activity that may require further investigation under financial crime and regulatory compliance processes.

Mastercard’s listing describes a role that combines AI application development with model evaluation, production engineering and monitoring. Rather than focusing exclusively on conventional statistical analysis or dashboards, the work includes building systems that retrieve relevant information, use tools, generate content and operate within technical and compliance constraints.

For candidates, this means that demonstrating an understanding of machine learning theory alone may not be enough. Practical evidence of building, evaluating and integrating AI applications will be especially relevant to the responsibilities described.

The work you may be responsible for

1. Building generative AI applications

The role involves implementing generative AI models for transaction-monitoring use cases. The listed technologies include LLMs, transformers and diffusion models, alongside systems capable of producing text, images or multimodal outputs.

Candidates should be prepared to explain how they have used AI models to address a concrete problem, what limitations they encountered and how they evaluated the resulting application.

2. Developing RAG and tool-based workflows

Mastercard specifically mentions:

• Retrieval-Augmented Generation (RAG)
• Model Context Protocol (MCP)
• Tool calling
• LangChain and LangGraph

These technologies support AI applications that retrieve external information, interact with tools and coordinate multi-step workflows. A strong portfolio project could demonstrate a complete workflow, from retrieving relevant documents to producing a grounded response and evaluating its reliability.

3. Fine-tuning and machine learning pipelines

The position includes developing pipelines for fine-tuning LLMs and other foundation models. This involves more than writing prompts: candidates should understand the model development process, how to assess performance and how to make repeatable improvements.

Python and frameworks such as PyTorch or TensorFlow are specifically identified in the job description.

4. Evaluating and monitoring AI systems

Mastercard highlights several areas that matter when AI applications move towards production:

• Detecting hallucinations and evaluating factual accuracy
• Checking for bias and scoring output quality
• Tracking latency, operating cost and model performance
• Applying model versioning, CI/CD and production monitoring
• Integrating models while considering performance, security and compliance

These requirements make evaluation and observability important parts of the role, not optional finishing touches.

5. Deploying AI with cloud and data platforms

The listing also names AWS, Azure and Databricks. Candidates should be ready to discuss how an AI application can be integrated into a larger software system, exposed through APIs and operated reliably.

The description does not identify a single mandatory cloud provider or specify a particular deployment architecture.

Skills to prioritise before applying

Use the following as a preparation guide based on the stated requirements—not as a separate official test syllabus.

Core technical areas

• Python and ML frameworks: Build and explain working solutions using Python, PyTorch or TensorFlow.
• LLMs and generative AI: Understand model capabilities, prompt engineering, fine-tuning and common failure modes.
• RAG and orchestration: Be able to describe retrieval, grounding, tool calling and multi-step workflows.
• AI evaluation: Explain how you would measure factuality, hallucinations, bias and output quality.
• Production practices: Understand APIs, monitoring, model versioning and ML CI/CD.
• Cloud and data tools: Gain practical exposure to AWS or Azure and, where possible, Databricks.
• Computer vision and multimodal AI: Be familiar with the image-analysis and multimodal capabilities mentioned in the listing.

Communication and problem-solving also matter. The role calls for collaboration with technical and non-technical stakeholders, so candidates should be able to explain technical trade-offs in clear language.

A useful portfolio project for this vacancy

If you are building a portfolio before applying, consider an AI-assisted document-review system using public or synthetic data. For example, it could retrieve relevant passages, use an LLM to produce a structured summary, cite supporting evidence and flag uncertain outputs for human review.

To make the project more convincing, document:

1. The problem and why a generative AI approach is appropriate.
2. The retrieval or tool-calling architecture.
3. The evaluation dataset and quality metrics.
4. Failure cases, hallucination checks and limitations.
5. API integration, logging, latency and cost considerations.
6. Security and privacy safeguards.

This would not replace relevant professional experience, and Mastercard has not stated that such a project is required. It is simply one way to demonstrate several of the technical capabilities named in the job description.

Who is eligible?

The stated educational requirement is a Bachelor’s or Master’s degree in Computer Science or an equivalent qualification.

The supplied listing does not specify a minimum CGPA, a particular graduation year, a fixed number of years of experience or whether fresh graduates are eligible. The title “Data Scientist I” should not be treated as confirmation that the role is exclusively for freshers.

Applicants should review the live vacancy for any additional eligibility conditions before submitting an application.

Salary, selection process and other missing details

Mastercard has not provided the following details in the supplied description:

• Salary or total compensation
• Minimum years of experience
• Application closing date
• Number of openings
• A confirmed interview or assessment sequence
• A specific work arrangement beyond the Pune location and full-time classification

Candidates should not rely on unverified salary estimates or assumed interview rounds when making application decisions.

APPLY NOW