Senior Associate AI And Data Science

National Payments Corporation Of India (NPCI); Mumbai Suburban; On-site; INR 2,000,000 - 3,200,000; Full time

Benefits offered by this job

Growth opportunities

Access to state-of-the-art compute

Proprietary datasets

Job summary

NPCI is seeking a hands-on AI/ML engineer to develop and deploy ML and GenAI models with a focus on fraud risk and secure banking interfaces. You will implement data and inference pipelines, optimize model serving on GPU clusters, and collaborate with senior architects to scale systems.

The role emphasizes RAG, fine-tuning, and guardrails to prevent hallucinations, with strong emphasis on distributed computing and high-throughput processing.

Qualifications

• Proficient Python programming and software engineering fundamentals including Git and testing.
• Hands-on ML model development using scikit-learn, XGBoost/LightGBM, and PyTorch or TensorFlow.

Responsibilities

• Build, train, and evaluate ML/GenAI models, moving from experimentation to production with guidance.
• Implement and maintain data and inference pipelines meeting reliability and latency targets.

Skills

Python

Git

Testing

scikit-learn

XGBoost

LightGBM

PyTorch

TensorFlow

LangChain

LangGraph

Hugging Face Transformers

SQL

Pandas

NumPy

Spark

Kafka

Flink

Docker

REST

gRPC

Kubernetes

MLflow

Cloud (AWS/GCP/Azure)

Statistics

Probability

Transformer

Tools

Kubeflow

Ray

DeepSpeed

Iceberg

Hudi

GPU/TPU usage

Docker

REST APIs

Kubernetes

Job description

Job Description:

Key Responsibilities:

• 1. Hands-On Execution (The "What")- Model Development: Build, train, and evaluate ML and GenAI models, taking them from experimentation to production with guidance from senior engineers.- Pipelines: Implement and maintain data and inference pipelines, ensuring they meet reliability and latency targets.- Collaboration: Work closely with senior architects and engineers, contributing to code reviews and learning production best practices.
• 2. Predictive AI & Scale (The "Fast")- Fraud at Scale: Contribute to our Fraud Risk Management (FRM) system, building features and models for streaming inference using Flink/Spark.- Feature Store: Help build and maintain features that serve both high-velocity transactional data and dense vector embeddings.- Optimization: Profile and optimize model serving to improve throughput on our GPU clusters.
• 3. Generative & Agentic AI (The "Smart")- AI Agents: Support the development of AI Agents that navigate banking interfaces, execute refunds, and detect social engineering attacks.- LLM Fine-Tuning: Assist in fine-tuning SLMs (Small Language Models) on secured financial data for edge deployment.- Safety Guardrails: Implement guardrails and evaluations that prevent hallucination in critical financial advice.

The Ideal Candidate Profile

You are a hands-on builder.

• You love to code, you are curious about how things work end-to-end, and you are eager to grow into a scale-veteran engineer.
• Essential Experience (3+ Years)- Dual Exposure: Hands-on experience in Classic ML AND/OR Modern GenAI (LLMs/Agents).
• Bonus if you have shipped either type of system to production.
• Scale Curiosity: Experience working with production systems and a strong desire to understand high-throughput, distributed computing.
• Deep Learning Fundamentals: Solid grasp of ML/DL fundamentals and the underlying math.
• Familiarity with Transformers or GNNs is a plus.
• Collaboration: Comfortable working in cross-functional teams and aligning your work with business goals.

Must-Have Skills-

• Programming: Strong proficiency in Python and solid software engineering fundamentals (clean code, version control with Git, testing).
• Core ML: Hands-on experience building, training, and evaluating ML models using scikit-learn, XGBoost/LightGBM, and PyTorch or TensorFlow.
• GenAI / LLMs: Practical experience with LLMs – prompt engineering, RAG, fine-tuning, and frameworks such as LangChain/LangGraph and Hugging FaceTransformers.
• Data Handling: Strong SQL and data wrangling skills (Pandas/NumPy); experience working with large datasets.
• Big Data / Pipelines: Working experience with Spark or Kafka/Flink for building data and inference pipelines.
• Deployment: Experience containerizing and deploying models (Docker), exposing them via REST/gRPC APIs, and basic familiarity with Kubernetes.- MLOps: Exposure to experiment tracking and model lifecycle tools (MLflow) and CI/CD for ML.
• Cloud: Hands-on with at least one major cloud platform (AWS, GCP, or Azure).
• Fundamentals: Solid grasp of statistics, probability, and deep learning concepts (including Transformer architecture).
• Communication: Ability to clearly document work and collaborate within cross-functional teams.

Technical Stack

• ML Ops: Kubernetes, Kubeflow, Ray, MLflow.
• Big Data: Spark, Kafka, Flink, Iceberg/Hudi.
• GenAI: vLLM, LangGraph, DeepSpeed, Unsloth.
• Cloud/Infra: Working knowledge of GPU/TPU usage and cost awareness.

The Package

• Growth: Learn directly from Principal Architects and senior engineers shaping the nations AI.
• Resources: Access to state-of-the-art compute infrastructure and proprietary datasets that are unique in the world.
• Legacy: The opportunity to say, "I helped build the AI that powers the Indian Economy.".

Requirements: