Juspay is a leading multinational payments technology company, powering superior conversion rates, seamless customer experiences, cost optimization, and fraud reduction at scale for 500+ top global enterprises and banks. Founded in 2012, the company processes over 300 million daily transactions, exceeding an annualized total payment volume (TPV) of $1 trillion with 99.999% reliability. Headquartered in Bangalore, India, Juspay’s global network of 1200+ payment experts operates across San Francisco, Dublin, São Paulo, and Singapore.
_
Juspay offers a modular and interoperable product suite for merchants that includes open-source payment orchestration, seamless authentication, payment tokenization, fraud & risk management, global payouts, end-to-end reconciliation, unified payment analytics, and more. For banks, Juspay’s offerings include end-to-end white label payment gateway solutions & real-time payments infrastructure._**
About the Role**

We are looking for a passionate and research-driven Machine Learning Engineer – NLP / LLM focused to take end-to-end ownership of building and deploying AI/ML systems, with a strong focus on Transformer based models. You’ll work at the intersection of large language models (LLMs), deep learning, and real-world product applications, helping us reduce hallucinations, enhance retrieval pipelines, fine-tuning models, and create smarter AI systems.

Key Responsibilities

  • Take end-to-end ownership of text understanding and RAG pipelines, from design to deployment.
  • Design, train, and evaluate LLM-based classifiers, prompt loops, and advanced NLP workflows.
  • Conduct research-driven development to improve factual accuracy and minimize hallucinations using structured data and retrieval methods.
  • Read papers and implement the SOTA techniques
  • Fine-tune and train transformer-based models, including domain-specific customizations.
  • Apply or quickly learn Reinforcement Learning (RL) techniques to optimize model behavior.
  • Work closely with cross-functional teams (engineering, product, data science) to integrate models into production systems.

Requirements

  • Proven experience with NLP techniques and deep learning frameworks, especially PyTorch.
  • Strong understanding of LLM internals, tokenization, embeddings, attention mechanisms, and transformer architectures.
  • Experience with prompt engineering, evaluation frameworks, and feedback loops.
  • Hands-on experience in fine-tuning, training custom models, and managing ML experiments.
  • Consistently learning Math to improve the fundamental understanding of Data Science and AI
  • Exposure to retrieval methods (vector stores, dense retrieval, hybrid search) and integrating them with generation models.
  • Familiarity with or willingness to learn Reinforcement Learning (e.g., RLHF).
  • Strong analytical and problem-solving skills with a passion for cutting-edge AI research.

Preferred Qualifications

  • Experience in deploying models in real-world production environments.
  • Trained models in Pytorch from scratch
  • Some ML Ops experience, deploying to GPUs
  • Background in Information Retrieval or Question Answering systems.
  • Publications, open-source contributions, or project portfolios in the NLP/LLM domain.
Apply Now