Use cases

Supported workflows in LiFE Suite.

Use cases describe what LiFE can support: documentation, AI data preparation, digital humanities research, classroom workflows, and reusable language resources. They are suggestions, not a list of completed projects.

01

Field linguists, students, and documentation teams

Language documentation and analysis pipeline

Start with preparing questionnaires then deploy this on Atekho for field data collection. Atekho allows for both field and follow-up remote data collection workflows. Atekho data is automatically synced to the cloud database for validation and further processing. Once data is validated and accepted, sentential and narrative data can be processed in MATra Lab with AI-assisted transcription, translation, and glossing. This could be used for analysing grammatical structures or writing grammatical descriptions of the language. Lexical data can be processed in LexiLab to produce multimodal, multilingual dictionaries. Finally, all outputs can be packaged for delivery to archives, for printing (especially dictionaries) or other downstream applications.

Recording and transcription projects Speaker and source metadata management

Supported work

Recording and transcription projects Speaker and source metadata management Lexicon and dictionary preparation Exports for analysis, archiving, and review
02

NLP teams, computational linguistics labs, and data teams

Data preparation for AI pipelines

You can start with recording data at scale on Atekho using its crowdsourcing and remote data collection workflows. Once data is collected, it can be processed in MATra Lab to produce AI-ready datasets for training models for speech recognition, machine translation, and other NLP tasks. MATra Lab supports AI-assisted transcription, translation, and custom annotation of all kinds of speech, text, image and video data. Scale up and accelerate your data preparation workflows by setting up multiple teams with 100s of members working in parallel, each with a fine-grained set of permissions, validating the work, managing permissions, and finally exporting the data in a format suitable for training AI models. You can even use Models Studio to train baseline models on your data and evaluate their performance.

ASR, OCR, MT, timestamps, glossing, and transliteration Review loops for model-assisted outputs

Supported work

ASR, OCR, MT, timestamps, glossing, and transliteration Review loops for model-assisted outputs Dataset packaging for downstream modelling Credit-based compute usage
03

Influencers, journalists, ethnographers, NGO workers, humanities researchers and social scientists

Digital Humanities and Social Science Research

If you are an influencer or a journalism or a researcher in the humanities or social sciences, you can use integrated AI models in MATra Lab to automatically transcribe and code multimodal and text data. You can combine your secondary materials (such as texts) with primary materials such as automatically transcribed interviews (of the authors or participants), focus group sessions and other research materials on a single platform. And then generate different kinds of visualisations and quantitative analyses describing your research.

Teams, permissions, and institutional contexts Project progress and contributor review

Supported work

Teams, permissions, and institutional contexts Project progress and contributor review Shared access to datasets and downloads Plan-based capacity for larger groups
04

Students, instructors, universities and training programmes

Classroom and field-methods projects

Use LiFE as a guided workspace for elicitation, transcription practice, lexicon building, and structured project submission. use the app to teach appropriate workflows and methods for social science and digital humanities research, field linguistics, language documentation, and computational linguistics. Instructors and TAs can set up projects for students to work on, review their work, and provide feedback. Students can work on their own or in teams, and submit their work for review. The app supports a variety of data types, including audio, video, text, and images, and provides tools for annotation, transcription, translation, and analysis.

Questionnaire-led elicitation Student project workspaces

Supported work

Questionnaire-led elicitation Student project workspaces Instructor review and feedback Reusable classroom datasets