A
IT(Software/Hardware)

Analyst-Data Science

American Express

Role Summary

American Express is hiring an Analyst-Data Science for its Analytics & Risk Management group in Gurugram, Haryana (Hybrid). In this hands-on engineering role, you will conduct model experimentation, inference optimization, evaluation, and deployment of modern AI systems. You will work across Large Language Models (LLMs), multimodal architectures, retrieval/reranking systems, and agentic workflows—profiling system bottlenecks, optimizing GPU memory, and turning cutting-edge ML research into scalable production components.

Key Responsibilities

  • Model Experimentation & Prototyping: Experiment with LLMs, multimodal models, embeddings, RAG frameworks, and agentic/tool-using systems in Python and PyTorch.
  • Inference Optimization: Run and optimize model inference runtimes using vLLM; manage continuous batching, KV caching, quantization, and throughput-latency trade-offs.
  • Evaluation & Quality Frameworks: Design evaluation frameworks to benchmark model behavior, output quality, and structured generation accuracy.
  • ML Systems Profiling & Debugging: Profile GPU utilization, memory usage, and latency bottlenecks; debug out-of-memory (OOM) failures and system issues.
  • Research & Implementation: Analyze AI research papers, reproduce state-of-the-art architectures, and translate promising ideas into production-quality software components.

Key Qualifications

  • Education: Bachelor’s or Master’s degree in Computer Science, Mathematics, Statistics, Engineering, Physics, or a quantitative discipline from a top institution.
  • Core Technical Stack: Advanced Python programming skills and working knowledge of PyTorch, transformers, tokenization, embeddings, attention mechanisms, and sampling methods.
  • Engineering Best Practices: Proficient in maintainable software development using Git, Docker, Linux, and REST APIs beyond Jupyter Notebooks.
  • ML Systems Expertise: Deep interest in model serving, vLLM runtimes, mixed precision, device placement, and hardware profiling (CUDA/Triton knowledge is a plus).
  • Demonstrated Depth: Background supported by substantial ML side projects, open-source contributions, research paper implementations, or inference/training benchmarks.

About American Express

American Express is a global financial services leader delivering credit card products, payment processing, and world-class customer experiences backed by a 175-year tradition of trust, service, and technical innovation.

Important Candidate Notice

This job posting and its descriptions are owned by American Express . We provide this overview to help developers discover opportunities. For applications, responses, or follow-ups, please connect with American Express directly. Job Central Hub does not process applications or review candidate queries.

Profile Match Score

Sign in to see how well your skills and experience align with this job.

Sign In
Overview
Expected Salary ₹12 - ₹22 Lakhs/Annum
Experience Required 0 - 3 Years
Last Date to Apply 26-09-2026 7 days left
Posted On 19-09-2026
Mandatory Skills
Python PyTorch Transformers LLMs Multimodal Models Model Inference vLLM KV Caching Continuous Batching Embeddings RAG & Vector Search Agentic Systems Docker Git Linux
Preferred Skills
Quantization Structured Generation GPU Profiling CUDA Triton Kubernetes NCCL Research Paper Implementation Evaluation Frameworks