GenAI & Agentic AI Application Engineer

About the Role

We are hiring a GenAI and Agentic AI Application Engineer to design build deploy and operate secure scalable enterprise AI applications across AWS and GCP. This is an application-engineering role focused on LLM integration Retrieval-Augmented Generation (RAG) AI agents local model hosting security evaluation and observability. Experience in banking financial services or another regulated industry is preferred.

What You Will Do

.

Build enterprise copilots knowledge assistants conversational applications document intelligence solutions and AI-enabled workflows.

.

Design production-grade RAG pipelines covering ingestion chunking embeddings vector and hybrid search reranking grounding citations and access-aware retrieval.

.

Develop single-agent and multi-agent solutions with tool calling workflow orchestration memory identity propagation human approvals and controlled execution boundaries.

.

Integrate foundation models through Amazon Bedrock Google Vertex AI approved model APIs and locally hosted open-source models.

.

Create secure APIs microservices asynchronous and event-driven workflows streaming responses structured outputs prompt management guardrails and fallback mechanisms.

.

Implement GenAI and agent evaluation for groundedness relevance hallucination citation accuracy safety task completion tool-selection accuracy latency and cost.

.

Establish end-to-end observability for prompts retrieval model calls agent actions tool calls token usage errors performance and infrastructure consumption.

.

Apply CI/CD automated testing infrastructure as code prompt and agent versioning controlled releases rollback and production support practices.

Required Experience

.

3 years of software or application engineering experience including hands-on delivery of GenAI applications to production.

.

Strong Python skills and experience with FastAPI Flask Django or equivalent backend frameworks.

.

Practical expertise in LLMs prompt engineering embeddings vector databases RAG structured outputs context management tool calling and AI agents.

.

Experience with agent or orchestration frameworks such as Amazon Bedrock Agents LangGraph LangChain LlamaIndex Semantic Kernel or equivalent.

.

Experience building REST APIs microservices asynchronous services event-driven applications and enterprise integrations.

.

Working knowledge of Docker Git Linux automated testing CI/CD and infrastructure as code.

.

Strong understanding of application security API security IAM encryption privacy controls and secure handling of sensitive enterprise data.

AWS Skills: Hands-on Experience Expected

.

Amazon Bedrock: foundation models Agents Knowledge Bases Guardrails and evaluation.

.

Amazon SageMaker: hosting and managing custom or open-source foundation models.

.

S3 Lambda EC2 and GPU instances ECS/EKS API Gateway DynamoDB RDS/Aurora PostgreSQL and OpenSearch.

.

Step Functions SQS SNS and EventBridge for workflow orchestration and asynchronous processing.

.

CloudWatch CloudTrail and X-Ray for monitoring audit logging and tracing.

.

IAM KMS Secrets Manager VPC and PrivateLink for identity encryption secrets and network isolation.

.

AWS CDK CloudFormation or Terraform for infrastructure as code.

GCP Experience

Experience with relevant GCP services including Vertex AI Gemini Model Garden Vertex AI Agent Builder Vertex AI Search Cloud Storage Cloud Run GKE Pub/Sub Workflows BigQuery Cloud SQL or AlloyDB Cloud Logging and Monitoring IAM Secret

Manager Cloud KMS and VPC Service Controls. Strong AWS expertise is required; practical GCP experience or demonstrated ability to build cloud-portable GenAI applications is expected.

Security Local Hosting & GenAIOps

.

Protect applications against prompt injection jailbreaks sensitive-data leakage unauthorized retrieval malicious documents insecure tool use and excessive agent autonomy.

.

Implement least privilege encryption private networking data masking or redaction access-aware retrieval tool allowlists validation rate limits human approvals and audit trails.

.

Host approved open-source models using SageMaker EC2 GPU ECS/EKS Vertex AI GKE or controlled private infrastructure; experience with vLLM Hugging Face TGI NVIDIA Triton or ONNX Runtime is desirable.

.

Optimize latency throughput concurrency GPU utilization context usage token consumption reliability and cost; implement autoscaling health checks load testing rollback and disaster recovery.

Preferred

.

Experience with multimodal AI document AI OCR intelligent document processing hybrid search knowledge graphs or reranking.

.

Knowledge of responsible AI AI governance model risk GenAI threat modelling OWASP guidance for LLM applications and regulated-industry controls.

.

Relevant AWS or GCP certification and ex