Agentic AI Architect

Job title

Program Manager

Location

Hybrid based out of Durham, NC

A Strategic Leader who can drive and deliver multiple AI Initiatives for Clients at account level as Leader/SPOC and FDE with hands-on expertise on the below areas:

• Design end-to-end AI/ML system architecture including data pipelines, model serving, inference infrastructure, and feedback loops
• Architect Retrieval-Augmented Generation (RAG) pipelines for knowledge retrieval from curriculum content, textbooks, and learning repositories
• Design multi-modal AI systems capable of processing text, audio, video, and assessments - Should have hands-on exp using Claude Code, Cursor, Blitzy Platforms
• Establish architecture for Adaptive Learning Engines — systems that personalize content difficulty, pacing, and style based on learner behavior
• Design LLM orchestration layers using frameworks such as Langchain, LangGraph, AutoGen, or Semantic Kernel
• Define model selection and routing strategies across providers (OpenAI, Anthropic Claude, Google Gemini, AWS Bedrock, Azure OpenAI, open-source Llama/Mistral)
• Architect AI tutor agents capable of multi-turn dialogue, Socratic questioning, hint generation, and step-by-step explanation
• Design autonomous assessment agents for auto-grading, rubric evaluation, plagiarism detection, and feedback generation
• Build content generation pipelines for lesson plans, quiz banks, summaries, and curriculum mapping using generative AI
• Define and govern the AI Harness — the full agentic AI stack from input processing to output safety and observability
• Define multi-agent orchestration patterns for collaborative AI tasks across curriculum design, student support, and institutional reporting
• Establish tool-use frameworks (MCP Protocol, OpenAI Function Calling) enabling agents to query databases, search content libraries, and trigger LMS workflows
• Define/Build Knowledge Graphs for EdTech domain mapping learning objectives, concepts, prerequisites, and competencies for structured curriculum reasoning
• Design learner memory systems tracking long-term progress, mastery levels, and knowledge gaps across sessions
• Implement LLM evaluation frameworks using RAGAS, DeepEval to measure faithfulness, relevance, and educational accuracy
• Define distributed tracing with Open Telemetry across the full AI request lifecycle for latency, cost, and quality monitoring
• Partner with instructional designers and curriculum experts to ensure AI systems align with pedagogical best practices
• Lead proof-of-concept and pilot programs for emerging AI capabilities (Adaptive learning system)
• Define and maintain the AI architecture roadmap, ensuring alignment with business goals and EdTech product strategy
• Should have experience working with neo4j, IRT based ML models, AWS AI services, Postgres, Edtech systems, AWS AgentCore, AWS Sagemaker and etc.

Mandatory skills

• AI Solution Architect
• Agentic AI
• AI/ML

Years of experience required

15-20