 Welcome to this brief introduction to On the Fly. I've earlier talked about how you can do real-time text mining, more specifically named entity recognition, on any web page using the extract tool. If you're not already familiar with extract, I strongly recommend that you go watch my brief introduction to that tool before continuing with this presentation. The most common question I get when talking about extract is does it work for PDF? And the answer to that question is unfortunately no. Because of the way extract is made, it can inherently only work on HTML pages. However, there's now a solution to that, and it's called On the Fly. On the Fly effectively allows you to run extract on most files and analyze the results coming out of it. More specifically, what can On the Fly do for you? On the Fly is a web resource accessible at this URL, which is linked down below. When you go to the website, you can upload your files, which could for example be PDFs of articles, have them converted to HTML, and then run text mining on them. This identifies the biomedical entities, such as genes and proteins and chemical compounds. It highlights them in the text like extract does, and allows you to retrieve molecular networks from the string and stitch databases and run enrichment analysis using both G-profiler and a GoJewel. How does On the Fly technically accomplish this? It uses several format conversion tools to deal with PDF files, spreadsheets, including Excel files, documents, including Word files, plain text files and even images using this setup. As you can see, we use one tool for PDF files, a different tool for the other file types, and finally image magic and optical character recognition to deal with bitmap images. The result is in either case an HTML page, which is subsequently sent to the extract server to obtain the annotated HTML page, as well as an entity list that can be sent to other APIs to retrieve networks and perform enrichment analysis. How can you use On the Fly? I see two main use cases. One is to simply extend the extract scope beyond webpages. Using On the Fly, you can do conversion of almost any file, keeping the layout of the page, and thus highlight entities, for example, in PDF files. The other use case is to use On the Fly to compile a team list from an article. You can use On the Fly to do automatic annotation of the article, considering both the full text, the figures and the tables. Here's an example of what happens when you run On the Fly on the figure inside an article. As you can see, most of the team names mentioned in this plot have been highlighted in red. You can then take the resulting team list and manually validate it to get the list for subsequent work. That's all I want to say about On the Fly. If you want to learn more about how you can use text mining in your work, I suggest you start looking at this presentation. Thanks for your attention.