<p><p><b>Description :</b></p><p><p><b><br/></b></p><br/>We are looking for an AI Engineer with strong expertise in Large Language Models (LLMs), Retrieval Augmented Generation (RAG), and AWS cloud infrastructure.
The candidate will architect, develop, and deploy AI agents and RAG pipelines, integrate vectorised knowledge bases, and build scalable applications leveraging Python, React, and containerised AWS services.<br/><br/><b>Responsibilities :</b></p><p><p><b><br/></b></p><br/>- Build and optimise RAG workflows with vector databases and embeddings.<br/><br/></p><p><br/>- Fine-tune, prompt-engineer, and deploy LLMs (Bedrock, GPT, Llama, Claude, etc.) for domain-specific production workloads.<br/><br/></p><p><br/>- Develop Python backend services (APIs, orchestration) and React UIs for AI-powered apps.<br/><br/></p><p><br/>- Deploy and manage AI workloads on AWS (EKS/ECS, Aurora Postgres, OpenSearch pgvector, SQS, Lambda, Elastic Cache, Secrets Manager).<br/><br/></p><p><br/>- Implement guardrails, monitoring, and evaluation for safe and reliable AI agents.<br/><br/></p><p><br/>- Collaborate with product and infra teams to ensure scalability, performance, and compliance.<br/><br/><b>Skills & Experience :</b></p><p><p><b><br/></b></p><br/>- 2- 4 years in AI/ML engineering with proven LLM + RAG project experience.<br/><br/></p><p><br/>- Strong in Python (AI pipelines, APIs) and exposure to React.js.<br/><br/></p><p><br/>- Hands-on with AWS AI & infra : EKS/ECS, Aurora PostgreSQL, OpenSearch/pgvector, SQS, Lambda, S3, Secrets Manager.<br/><br/></p><p><br/>- Knowledge of databases: Postgres, MongoDB, Neo4j, Vector DB.<br/><br/></p><p><br/>- Familiar with Lang-Chain, Llama-Index and Hugging Face.<br/><br/></p><p><br/>- Good understanding of containerization (Docker/Kubernetes) and CI/CD Knowledge of async processing, DLQ handling, and scalable message-driven architectures.<br/><br/><b>Must to Have :</b></p><p><p><b><br/></b></p><p>- Experience with Amazon Bedrock and building LLM-based agents.</p><p><br/></p><br/>- ML-Ops exposure (ML flow, W&B).<br/><br/></p><p><br/>- Graph/RAG hybrid architectures (Neo4j + vector DB).<br/><br/></p><p><br/>- Experience with prompt guardrails, structured output validators, and secure data handling.<br/></p><br/></p> (ref:hirist.tech)