REF89610I-Senior Program Analyst - Python, GEN AI, Agentic AI, 2+ years experience

WNS Global Services

About The Position

WNS (Holdings) Limited (NYSE: WNS), is a leading Business Process Management (BPM) company. We combine our deep industry knowledge with technology and analytics expertise to co-create innovative, digital-led transformational solutions with clients across 10 industries. We enable businesses in Travel, Insurance, Banking and Financial Services, Manufacturing, Retail and Consumer Packaged Goods, Shipping and Logistics, Healthcare, and Utilities to re-imagine their digital future and transform their outcomes with operational excellence.We deliver an entire spectrum of BPM services in finance and accounting, procurement, customer interaction services and human resources leveraging collaborative models that are tailored to address the unique business challenges of each client. We co-create and execute the future vision of 400+ clients with the help of our 44,000+ employees. Job Description · Develop agentic AI workflows using LangChain/LangGraph with custom tools, memory, decision logic, and MCP integrations. · Implement LLM applications with advanced prompt engineering (structured outputs, function calling, context management) and build RAG pipelines with vector databases. · Build and deploy event-driven AI services on AWS using Bedrock, SageMaker, Lambda, EventBridge, Step Functions, API Gateway, DynamoDB, and S3. · Write clean, testable Python code; create unit/integration tests; and package solutions with CI/CD pipelines (GitHub Actions). · Monitor and optimize AI application performance through logging, metrics, token usage, and cost management. · Integrate AI services with APIs, backend systems, and databases to ensure secure and reliable data flow. · Contribute to document AI and predictive analytics models for information extraction, classification, and forecasting. · Fine-tune Hugging Face Transformer models (BERT, RoBERTa, LayoutLM, DistilBERT) and train neural networks (PyTorch/TensorFlow) for supervised tasks. · Implement data preprocessing and annotation workflows for training and evaluation. · Work with OCR and text-processing tools (AWS Textract, Tesseract) for semi-structured/unstructured documents.

Requirements

5+ years in software engineering, including 3+ years in python, AI/ML or Generative AI systems.
Proven development experience with LLM frameworks (LangChain, LangGraph) and LLM APIs (Claude, OpenAI, Llama, Titan).
Hands-on experience with AWS Bedrock and SageMaker for AI/ML workloads.
Strong Python skills (async, data structures, optimization).
Practical experience with vector databases (Pinecone, ChromaDB, FAISS, OpenSearch) for RAG implementations.
Familiarity with event-driven AWS services (Lambda, EventBridge, Step Functions, API Gateway, DynamoDB, S3, CloudWatch).
Exposure to document AI and predictive analytics, including Hugging Face Transformers, classification, and extraction pipelines.
Understanding of data preprocessing and OCR workflows for document pipelines.

Responsibilities

Develop agentic AI workflows using LangChain/LangGraph with custom tools, memory, decision logic, and MCP integrations.
Implement LLM applications with advanced prompt engineering (structured outputs, function calling, context management) and build RAG pipelines with vector databases.
Build and deploy event-driven AI services on AWS using Bedrock, SageMaker, Lambda, EventBridge, Step Functions, API Gateway, DynamoDB, and S3.
Write clean, testable Python code; create unit/integration tests; and package solutions with CI/CD pipelines (GitHub Actions).
Monitor and optimize AI application performance through logging, metrics, token usage, and cost management.
Integrate AI services with APIs, backend systems, and databases to ensure secure and reliable data flow.
Contribute to document AI and predictive analytics models for information extraction, classification, and forecasting.
Fine-tune Hugging Face Transformer models (BERT, RoBERTa, LayoutLM, DistilBERT) and train neural networks (PyTorch/TensorFlow) for supervised tasks.
Implement data preprocessing and annotation workflows for training and evaluation.
Work with OCR and text-processing tools (AWS Textract, Tesseract) for semi-structured/unstructured documents.