GenAI Solution Architect

Job description

**Role Overview** - This role sits at the intersection of traditional software engineering and machine learning. The primary goal is to solve complex enterprise problems by securely integrating AI endpoints into existing systems. **Key Responsibilities** - Agentic AI & RAG: Design multi-step workflows, autonomous AI agents, and retrieval pipelines connecting models to proprietary, real-time data. - Model Fine-Tuning & Prompt Engineering: Optimize existing foundational models using Parameter-Efficient Fine-Tuning (PEFT) like LoRA/QLoRA and systematically refine system prompts. - Evaluation & Guardrails: Implement robust validation, golden dataset evaluations, and output guardrails to manage hallucinations and ensure safe, compliant responses. - Deployment: Containerize models and integrate them as REST APIs into production environments. **Must-Have Skills & Tech Stack** - Programming Languages: Advanced proficiency in Python. - Orchestration Frameworks: LangChain, LangGraph, LlamaIndex, or Hugging Face. - Vector Databases: Pinecone, ChromaDB, Weaviate, or pgvector. - Cloud AI Platforms: Amazon Web Services (AWS Bedrock), Microsoft Azure (Azure OpenAI), or Google Cloud (Vertex AI). - Concepts: Transformer architectures, tokenization, model context protocol (MCP), and MLOps.