Automating Technical Writing: How to Get Started using R and Python

We describe the secrets to our success at an efficient project we recently carried out to automate the initial steps of report writing. Our experiences with project management, communication plans, tool iterations and validation will be reviewed. The key details of how the current version of the tool operates will be shared. In brief, a user provides a template (.txt) and the raw data files (.csv) to the tool, which results in a Microsoft Word report first draft of all tables in the report in ~30 seconds. The key concepts to successful use of R and Python in automating portions of writing will be described and fleshed out in context of the development process. Next steps are explored including leveraging the LLM models to further train the tool to automate additional steps in report writing. This tool (and tools like it) can be expected to save ~30 to 40 hours per report. Later iterations using LLM models are discussed which have the potential to improve technical writing productivity multifold thereby accelerating drug development timelines.