AI Software Engineer III

Position Summary The AI Software Engineer III will be responsible for designing, developing, and supporting secure, scalable AI-enabled applications. The role focuses on Python development, Generative AI, prompt engineering, Retrieval-Augmented Generation (RAG), document intelligence, AI agents, and enterprise AI integration. Location: Farmington Hills, MI (Onsite – 4 days per week) Experience: 8+ Years Work Arrangement: Onsite – Local candidates strongly preferred Key Responsibilities - Design and develop production-quality applications and AI services using Python. - Build Generative AI capabilities using Amazon Bedrock and Large Language Models (LLMs). - Design, develop, test, and optimize prompts, including prompt templates, structured outputs, and reusable prompt patterns. - Develop RAG, intelligent search, embedding, and vector retrieval solutions. - Build AI agents and agent-based workflows using frameworks such as Strands Agents. - Develop reusable AI services and APIs using FastAPI and Pydantic. - Integrate AI solutions with structured and unstructured enterprise data. - Support document processing and extraction using Amazon Textract. - Implement vector-based retrieval using PostgreSQL/pgvector. - Establish safeguards for AI output quality, reliability, validation, traceability, and secure data handling. - Optimize AI solutions for performance, latency, token usage, and cost. - Develop automated testing and evaluation capabilities for AI solutions. - Collaborate with application, data, DevOps, QA, architecture, cybersecurity, and business teams. Required Qualifications - 8+ years of experience in Python development, software engineering, AI/ML engineering, or related fields. - Strong hands-on experience with Python, REST APIs, FastAPI, Pydantic, and boto3. - Experience with Amazon Bedrock, LLM integration, and Generative AI solutions. - Strong expertise in prompt engineering, including prompt design, testing, optimization, structured outputs, and evaluation. - Experience with RAG, embeddings, semantic search, vector databases, and PostgreSQL/pgvector. - Experience developing AI agents and agentic workflows, preferably using Strands Agents. - Hands-on experience with Amazon Textract, AWS, Docker, and automated testing. - Strong understanding of AI security, data protection, responsible AI, and enterprise governance. - Experience working in large enterprise or regulated environments. Interview Process - Internal Technical Screening: 20–30 minutes. - Client Interview: Virtual.