NeuralIR is applying AI to scientific information retrieval (IR). New methods are needed to deal with the exponential growth in content. NeuralIR's solutions integrate search directly in cloud notebooks (notebook.studio), allowing more productive scientific workflows. The vast range of open access research is organized using an open experimental mathematical object index (xmoi.org). By spanning two dynamic online communities, the Python/DataScience ecosystem of notebooks and open access research, NeuralIR can scale-up support for computer-aided workflows in science, technology, economics, finance, reporting and government.