Join Arvexi to build AI that reads lease documents better than humans. We're an AI-powered lease accounting platform replacing the $2.5B manual abstraction industry.\n\nAs our AI/ML Engineer, you'll work on document intelligence — building models that extract 50+ fields from lease PDFs with confidence scoring, classify document types, and detect embedded leases. You'll own the extraction pipeline end-to-end.\n\nWhat you'll do:\n- Build and improve AI extraction models for lease document understanding\n- Design confidence scoring systems for financial data accuracy\n- Develop document classification and entity recognition pipelines\n- Work with real enterprise lease data (not toy datasets)\n- Ship to production — your models serve paying customers\n\nWhat we're looking for:\n- 3+ years ML/AI engineering experience\n- Strong Python, PyTorch or TensorFlow\n- Experience with NLP, document AI, or information extraction\n- Comfort with production ML systems (not just notebooks)\n- Bonus: experience with financial documents, OCR, or layout-aware models\n\nWhy Arvexi:\n- Founded by ex-OpenAI and Google engineers\n- Real AI solving a real problem (not another chatbot wrapper)\n- Early-stage — massive impact on product direction\n- Competitive equity + salary\n- Remote-first, async cultureOriginally posted on Himalayas