Ai Architect

About the Role

We are looking for a hands-on AI Architect to design, build, and deploy production-grade Generative AI systems on AWS. This role goes beyond experimentation—you will architect secure, scalable, and costefficient GenAI solutions used by real users in enterprise environments. You will work closely with engineering, data, and product teams to deliver LLM-powered applications, including RAG-based document intelligence, chatbots, and AI assistants.

Key Responsibilities

• Architect and implement Generative AI solutions using LLMs (GPT, Claude, Mixtral, etc.)

• Design and deploy Retrieval-Augmented Generation (RAG) pipelines for document Q&A and enterprise search

• Build semantic search and embedding pipelines using vector databases (FAISS, OpenSearch, Pinecone)

• Select and optimize LLM models, prompts, and inference strategies for accuracy, latency, and cost

• Implement hallucination mitigation techniques (grounding, prompt constraints, validation layers)

• Design secure, scalable architectures on AWS (Bedrock, SageMaker, Lambda, API Gateway, S3)

• Fine-tune models using PEFT techniques (LoRA, QLoRA) when required

• Partner with MLOps teams to productionize models with CI/CD, monitoring, and rollback

• Optimize GenAI systems for cost, latency, and throughput

• Collaborate onsite with cross-functional teams (3 days/week in Raleigh)

Required Skills & Experience

Generative AI & LLMs

• Strong understanding of LLM architectures and inference

• Hands-on experience with RAG systems in production

• Prompt engineering, temperature/top-p tuning

• Knowledge of LoRA / QLoRA / PEFT techniques

• Experience mitigating hallucinations and improving factuality

Embeddings & Retrieval

• Semantic embeddings (Sentence-BERT, OpenAI, etc.)

• Chunking strategies and metadata handling

• Vector similarity search (cosine, dot-product)

• Vector databases: FAISS, OpenSearch, Pinecone

AWS & Cloud Architecture

• AWS AI/ML services: Bedrock, SageMaker

• Serverless & APIs: Lambda, API Gateway

• Data storage: S3, DynamoDB

• Security: IAM, KMS, VPC, CloudTrail

• Experience designing enterprise-grade, compliant systems

Programming & Frameworks

• Python (strong)

• Experience with LangChain, Haystack, FastAPI (or similar)

• Familiarity with async processing and caching layers

MLOps & Production

• Model versioning and monitoring

• CI/CD for ML systems

• Rollback strategies and drift detection

• Performance and cost monitoring

Nice to Have

• Experience with knowledge graphs integrated into GenAI

• PDF/document ingestion pipelines (OCR, Textract)

• Multi-tenant GenAI architectures

• Healthcare / Pharma / regulated industry experience

• Exposure to self-hosted open-source LLMs

Qualifications

• Bachelor’s or Master’s degree in Computer Science, AI/ML, or related field

• 7+ years in software/ML engineering, with 2+ years in GenAI/LLMs

• Proven experience deploying AI systems to production