About the Role

Invisible Mechanics is seeking a Founding AI/ML Engineer to join their remote team in India. This is a unique opportunity to contribute to an AI-powered adaptive learning platform designed for students preparing for India's highly competitive JEE & NEET examinations. Unlike conventional roles, this position focuses on deep AI application, leveraging a proprietary taxonomy with 54 metadata fields and 1,550+ tags to precisely diagnose student misconceptions and craft personalized learning paths.

The core of this role involves developing sophisticated AI systems, not merely wrapping existing large language models. The AI drives two critical functions: (1) automated, scalable content tagging for administrators using Claude API and Mathpix, and (2) student-facing adaptive sequencing, detailed misconception diagnosis, and prerequisite tracing from structured data. All student-facing content is meticulously pre-authored, ensuring a consistent and high-quality educational experience.

What You'll Do

• Design and iterate on the core adaptive learning algorithm to personalize student learning paths based on diagnosed knowledge gaps.
• Build, optimize, and scale the content ingestion pipeline, utilizing Claude Sonnet/Haiku/Opus and Mathpix to process over 1,400 problems monthly.
• Develop the robust misconception diagnosis engine and the prerequisite chain tracing logic that informs adaptive sequencing.
• Design efficient backend APIs and data schemas to support a complex 54-field taxonomy at scale.
• Lead prompt engineering efforts specifically for structured data extraction pipelines, moving beyond general chatbot applications.

What We're Looking For

• Experience: 2-5 years of hands-on experience in building backend systems, data pipelines, or impactful machine learning products.
• Technical Skills: Strong proficiency in Python and/or Node.js.
• LLM Expertise: Demonstrated real-world experience with LLM APIs, including structured extraction, advanced prompt engineering, output parsing, and cost optimisation.
• Database Skills: Solid understanding of Postgres and database design principles for managing complex relational data.
• Mindset: A strong systems thinking approach, understanding how architectural decisions influence student outcomes and product performance.

Bonus Points

• Prior experience with Mathpix or OCR technologies.
• Familiarity with educational psychometrics, specifically Item Response Theory or Bayesian Knowledge Tracing.
• Experience with knowledge graphs.
• A background in EdTech (educational technology).
• Expertise in scientific content processing.

What We Offer

• Significant founding team equity (ESOP).
• A competitive monthly salary of ₹1.5L - ₹2L, with 6-monthly performance-based raises.
• The opportunity to tackle one of the most intellectually stimulating AI challenges in the Indian EdTech sector—focusing on deep diagnostic problems like "diagnose why this student keeps failing rotational dynamics" rather than simple summarization tasks.
• Work closely with a founder who personally built the core taxonomy and Claude pipeline.

How to Apply

To apply, please share a project where you took the lead in designing a complete system, not just an isolated model. We value speed and competence. Additionally, include a brief note explaining why this specific project at Invisible Mechanics excites you. We are looking for candidates whose vision aligns with ours. Expected joining period: April–May 2026.