Jeremy Howard

AI Infrastructure Architect

Sources checked
ABOUT

Co-founder of fast.ai and CEO of Answer.ai. Former President and Chief Scientist at Kaggle. Co-developed ULMFiT (Universal Language Model Fine-tuning), which established modern transfer learning in NLP before BERT and GPT. Author of 'Deep Learning for Coders with fastai and PyTorch'. Currently building GPU-accelerated agentic development environments and open foundation tools.

Areas of focus

Professional niches

THE WORK BEHIND THE PROFILE

Proof of Work

implementationChecked Sep 20, 2026

Fastai Deep Learning Framework and Layered API

A layered high-level and low-level deep learning library built on top of PyTorch that drastically democratizes state-of-the-art computer vision, NLP, and tabular models with decoupled optimizer and scheduler primitives.

Scope & limitations

Designed primarily for PyTorch ecosystem; high-level abstractions require delving into mid-level API for highly custom multi-modal tensor graphs.

Context: PyTorch 2.4, CUDA 12, mixed-precision FP16/BF16 training, One-Cycle learning rate policies.

View mission
researchChecked Sep 20, 2026

ULMFiT: Universal Language Model Fine-tuning for Text Classification

Seminal research paper introducing inductive transfer learning to natural language processing via 3-step pre-training, discriminative fine-tuning, and slanted triangular learning rates, outperforming prior custom architectures by orders of magnitude.

Scope & limitations

Evaluated on recurrent AWD-LSTM architectures prior to pervasive multi-head self-attention scale; inductive bias differs from current masked and autoregressive transformers.

Context: AWD-LSTM 3-layer language model pre-trained on WikiText-103; foundation for modern BERT/transformer transfer learning pipelines.

View mission

Guides to evaluating AI expertise