AirflowAWSDockerPythonPyTorchTensorflowAIMachine LearningMLDeep LearningLLMLarge Language ModelsOpenAIGeminiAgenticTensorFlowMLOpsMLflowECSEKSLambdaGlueSageMakerBedrockKinesisCodePipelineGitVersion ControlAgileCI/CD
About this role
Role Overview
Design, build, and deploy production-ready AI/ML systems on AWS with focus on reliability, scalability, and performance for banking applications
Implement and maintain MLOps pipelines using AWS services (SageMaker, Bedrock, Lambda, Step Functions) including model versioning, monitoring, and automated retraining workflows
Build and optimize AI solutions using AWS Bedrock, OpenAI API, and Gemini API combining with Model Context Protocol (MCP), Agent-to-Agent (A2A) protocol for various banking use cases
Design and implement prompt engineering frameworks and prompt management systems for LLM-based applications
Develop graph analysis solutions for fraud detection, customer relationship mapping, and network analysis in banking contexts
Debug and troubleshoot production AI systems, identifying and resolving issues in model performance, data pipelines, and AWS infrastructure
Build and maintain AIOps practices including automated monitoring, alerting, and incident response for AI systems on AWS
Optimize model serving infrastructure for latency, throughput, and cost-efficiency using AWS services
Implement robust data pipelines using AWS Glue, Kinesis, and related services for training and inference
Collaborate with software engineering and risk teams to integrate AI capabilities into banking products and services
Ensure compliance with banking regulations and security standards in all AI deployments
Monitor model performance in production and implement drift detection and retraining strategies
Stay current with latest AI research papers and breakthroughs, evaluating applicability to banking and financial services
Research and prototype emerging AI architectures and techniques for financial use cases
Evaluate new paradigms in model training, inference optimization, and architectural innovations
Share knowledge through technical discussions, paper reviews, and internal research presentations
Identify opportunities to apply cutting-edge research to improve fraud detection, customer service, risk assessment, and other banking operations
Requirements
Master's or PhD in Computer Science, AI/ML, Mathematics, Statistics, or related field with focus on machine learning, deep learning, or strong publication record or demonstrated deep understanding of AI research (thesis, projects, or contributions to the field)
Deep theoretical knowledge of modern AI architectures and training methodologies
Deep understanding of transformer architectures and attention mechanisms
Strong knowledge of large language models (LLMs), multimodal models, and their architectural evolution
Familiarity with current research trends including Agentic AI systems
Experience integrating and managing external AI APIs: OpenAI API (GPT models, embeddings, fine-tuning), Gemini API (Google's multimodal models)
Expertise in prompt engineering, prompt management, and LLM-powered Agents orchestration frameworks
Strong knowledge of graph databases and graph analysis techniques: AWS Neptune or similar graph databases
Strong programming skills in Python and experience with ML frameworks (PyTorch, TensorFlow)
Hands-on experience with MLOps tools (MLflow, Weights & Biases, Airflow)
Experience with containerization and orchestration (Docker, ECS, EKS)
Strong understanding of distributed training and GPU optimization on AWS
Experience with CI/CD pipelines using AWS CodePipeline or similar
Strong software engineering principles and version control (Git)
Experience in the banking or finance industry is a plus.