Artificial Intelligence (AI) Engineer – Backend Focus

Position Summary:

A backend-focused AI engineer responsible for developing secure, scalable, and production-grade AI applications, with deep experience in LLM integration, retrieval-augmented generation (RAG) pipelines, including Graph-RAG, Agentic AI, and cloud-based LLM Ops workflows. The role emphasizes Amazon SageMaker Studio, ECS, ECR, lambdas, Agentic Core, APIs, OpenSearch Vector DB, and Dynamo DB for operationalizing GenAI-powered Digital Products within FedRAMP-compliant AWS environments.

Responsibilities:

AI Solution Development:

Expert hands-on building of RAG and Graph-RAG architectures to handle multiple complex data formats (PDF, images, tables, Word documents, Excel, acronyms, attachments, etc.) to create cleansed standardized data for hydration into a vector database.

Expert hands-on knowledge on text embeddings, image embeddings, chunking logic, metadata creation, and embedding vectors indexing.

Expert hands-on knowledge in creating a highly accurate RAG retrieval system with knowledge on reranking, semantic search, similarity search, hybrid search, etc., to search by text or images.

Implement secure, scalable, highly accurate RAG, Agentic AI pipelines using LangChain, Strands, MCP, A2A frameworks, or AWS-native services like Bedrock, Agentic Core, OpenSearch Vector Database, and Knowledgebase.

Create backend infrastructure for chatbot applications with long-term and short-term memory capabilities to improve user experience.

Hands-on knowledge of creating APIs, Graph-RAG, develop agentic AI workflows with MCPs, A2A, and Skills.

AI/ML Skills:

Experience operationalizing AI/ML pipelines in SageMaker Studio with model governance

Experience with Amazon - Bedrock, Agentic Core, OpenSearch Vector Database, knowledgebase, lambda, API Gateway, FASTAPIs or Flask, SQS, SNS, Step functions, DynamoDB, RDS/Postgres SQL, ECS, ECR, IAM, CloudWatch, and EKS or Fargate.

Frameworks: LangChain, LangFuse, LlamaIndex, Strands, RAGAS, CrewAI, MCP, and A2A.

Prompt engineering, LLM evaluation methodologies, bias detection, and hallucination detection.

LLM Integration & LLM Ops:

Integrate multiple LLMs via APIs (AWS Bedrock: Anthropic - Claude, Titan, Llama, Stability Diffusion models)

Implement structured prompt engineering frameworks, response evaluation tools, and feedback loops

Build model optimization layers, including prompt selectors, model switchers, and cache layers

Cloud Infrastructure & Deployment:

Deploy AI services using SageMaker, ECS, Lambdas, Agentic Core, and Elastic Load Balancers

Containerize backend systems with Docker and deploy to scalable environments using ECS/EKS

Implement CI/CD pipelines via GitHub Actions integrated with AWS Systems Manager and CodePipeline

Architect solutions for VPC isolation, IAM hardening, and FedRAMP High compliance

System Integration & Maintenance:

Integrate AI workflows with enterprise databases, legacy platforms, and identity providers

Monitor service performance, GPU utilization, and system health via CloudWatch and custom logging

Build automated testing pipelines for model accuracy, bias detection, and system robustness

Maintain technical documentation and developer runbooks for long-term system support

Work Environment:

Remote-first with collaborative engagements and occasional client travel

Mission-focused development aligned with executive priorities

Continuous learning and rapid prototyping of cutting-edge AI technologies

Agile delivery culture with strong cross-functional collaboration

Required:

Required / Minimum Qualifications

10+ years of IT experience.

3+ years of experience as an AI Engineer

3+ years of experience in AWS

AWS Services: Graph RAG, Bedrock Agentic Core, EC2 (GPU-enabled), SageMaker (Studio, Pipelines, Endpoints, Model Registry), Bedrock, OpenSearch Vector DB, Systems Manager, Load Balancers, Amazon - Bedrock, OpenSearch Vector Database, knowledgebase, lambda, API Gateway, FASTAPIs or Flask, SQS, SNS, Step functions, DynamoDB, RDS/Postgres SQL, and EKS or Fargate

Proficient in coding: Python (async, FastAPI, LangChain, Transformers) and Terraform

DevSecOps: Docker, GitHub, GitHub Actions, CI/CD pipelines

Cloud-Native Development: Infrastructure-as-Code, cloud monitoring, and security policies

Preferred:

Preferred / Nice-to-Have Qualifications

Experience with React or other frontend frameworks for full-stack AI interfaces (Streamlit, ReactJS, JavaScript, Typescript, HTML, and CSS)

Government/federal sector AI solution experience with FedRAMP High or FISMA

Bachelor''s or equivalent in Computer Science, Software Engineering, AI/ML, or related technical field

AWS certifications (Machine Learning Specialty, Solutions Architect) a strong plus

Experience using AI coding assistant tools like OpenAI Codex and Claude Code.