• A research team is asked to justify their decision-making and analysis from qualitative data for a work submitted for publication.  They want to attain some machine back-up in their analysis (originally from human close-reading and manual coding).  They decide to use NVivo 12 Plus to model topics from text-based survey responses to justify their extracted themes and subthemes.  
• Students are conducting various types of literature reviews for different projects. They use NVivo to sift through mounds of machine-readable .pdf files to identify the main themes of the respective works, so they know which ones to read in-depth.  They use their findings for research leads, research design, analytical papers, and other applications.  
• A research team dealing with a social topic is interested in some of the public debates going on on social media. They use NCapture (of NVivo) to capture data from several related microblogging sites, and from a major social networking site.  They use NVivo to analyze the intercommunications, the social networks, and the sentiment around particular hot-button topics.  They use this information in an augmentary way with other research.  
• A researcher has been conducting a multi-year project.  She has been evolving a codebook on this topic for years. This codebook is informed by top-down coding (from models, theories, frameworks, research questions)...and bottom-up coding ("grounded theory").  By the way she has ingested data into the project, she is able to query the data in very particular ways and output particular reports that answer very specific questions, including those that she had not conceptualized prior.  
• A research team has been seated around the world.  The team members have all trained on a particular consensus-created codebook and are now busy coding particular policies related to their own countries based on this codebook.  They are all aiming for a Kappa coefficient of .6 to .7, the range which they had achieved in the training based on the codebook and a particular set of data.  Now, they are coding different policies, and they code on their own projects, and their projects are collated into a master project file for analytics.  
• A research team has co-created a codebook that they want others to be able to inherit and use in NVivo 12 Plus.  They output their codebook in the Codebook Exchange Format...for interoperability across multiple projects and softwares.  

• the ingestion of multimedia data (text, imagery, audio, video, and other file types) into a project for curation and management;
  
• the capability to annotate (memo) all ingested files and built-in-NVivo files;
  
• the maintenance of research journals (in close proximity to the data);
  
• the ability to “query” the structured and unstructured data in a variety of ways (word frequency counts, text searches, coding queries, matrix queries, and others); 
• the ability to “autocode” data based on either structure or on human-based code patterning (human-supervised machine learning);
• the ability to "autocode" based on theme and sub-theme extraction (unsupervised topic modeling);
• the ability to "autocode" based on sentiments (extracted from a pre-labeled built-in sentiment set); 
• the ability to integrate the work of a team co-coding a dataset or mixed datasets;
• the ability to set up and run fine-tuned matrix and cross-tabulation table queries; 
• the ability to auto-create related data visualizations (from data queries as well as ingested data);
• the ability to depict relationships between nodes in a project and draw 2D node-link models; 
• the ability to auto-create and output reports;
• the ability to archive and search through a variety of multimedia files, 
• the ability to output codebooks as Word files and as .qdc files, 
• and many others.

These functions may be used in whole or in part.  What data is collected, how it is processed, how it is queried, and such, all affect what may be learned from the tool and asserted from the data.