AI Engineering Intern

About the Role

You’ll work directly with school teachers who are learning to build their own apps with Claude. They start with an idea and often a working prototype. Your job is to mentor them on what AI can do, help them get unstuck, and then turn their half-finished apps into polished, reliable products that real students and teachers use every day. This role is for someone who is just as comfortable explaining an idea patiently to a teacher as they are shipping production code.

Key Responsibilities

• Work closely with teachers: understand their classroom problems, coach them on using Claude and AI tools, and help them build confidence as creators
• Take teacher-built prototypes and turn them into production-ready apps with clean architecture, testing, deployment and good UX
• Design and build end-to-end agentic applications: tool use, multi-step workflows, MCP integrations and evaluation
• Build data extraction pipelines that turn textbooks, syllabi, PDFs and scanned material (including regional-language content) into structured data
• Build RAG systems over curriculum content so answers are accurate and grounded
• Model subjects as knowledge graphs that link concepts, prerequisites and learning outcomes across grades
• Build interactive lab simulations for topics across physics, chemistry, biology, math and other subjects, so students can learn by doing even where there are no physical labs
• Write simple documentation and guides so teachers can keep using and extending what you build

Requirements & Tech Stack

• Basic programming skills in Python and/or TypeScript/JavaScript, with strong working knowledge of Claude Code and Cowork
• Hands-on experience building LLM applications: agents, tool calling, prompt design, and working with APIs like Claude or OpenAI
• Experience with RAG: chunking, embeddings, vector databases, retrieval evaluation
• Experience with data extraction from unstructured documents (PDF parsing, OCR, structured outputs)
• Familiarity with knowledge graphs (Neo4j or similar, ontology design, or graph-based retrieval)
• Ability to build interactive web front-ends and visual simulations (React, Canvas/SVG, p5.js, Three.js or similar)
• Experience taking projects all the way to deployment, not just notebooks
• Clear, patient communication. You can explain technical ideas in simple words.

Qualifications

• A real interest in education and social impact. You care whether a student in a village school understands the concept.

Nice to have:

• Ability to speak Telugu or another Indian regional language
• Teaching, tutoring or volunteering experience
• Experience with the Claude Agent SDK or building MCP servers
• A portfolio of shipped projects, open-source work or hackathon builds