TRAI-L is a hands-on laboratory. We believe in the agile paradigm to help bring technology to niches where commercial enterprises  have little or no interest, but that are, nonetheless, key to helP solve social issues.

All of our projects are subject of transdisciplinary research, applied research, and augmented innovation leading to the development and deployment of solutions on the real world. Still, research and production are not dependant on each other, rather we believe thay are complementary. Research can improve the applications and applications can bring back data for research.
We host two types of projects:
In-lab projects. projects we generate and nurture inside the lab.
Supported projects. projects that started outside the lab and  align with our social scope, to which we provide knowledge, research, connections, exposure and hands-on development.

In-lab projects.


The Association for the Advancement of Minorities through Natural Language Processing ( is a nascent organization and the largest working project of TRAI-L that seeks to use Natural language Processing Techniques for the improvement of the lifes of minorities. It encompasses not only researchers or academics but also activists in favor of social causes. A large body of volunteers works within AMNLP supporting many of the other projects of this laboratory, namely: Voisell, Pioli, Indlex antd external project Miyotl.


Mobile application that allows easy access to online buying and selling markets for people with visual disabilities, by providing an audio-haptic interface mediated by voice recognition and synthesis algorithms, and that has robustness in interaction by integrating algorithms of Natural Language Understanding (NLU). In this way, in conjunction with a touch screen interface friendly to populations with vision loss, the application can be used with great naturalness and intuitiveness through audio instructions and thus offer a viable alternative to people with visual disabilities.


The new concept of social networking for the interaction of minority language speakers. Currently supporting registration for 68 mexican originary languages and 348 variants. Pioli incorporates a range of possibilities that will allow users to choose from more than 12 emotions and reactions, turning a simple “Like” into more than just a word and providing posts or “Kuikats” (Náhuatl for trill) with the right tone and context.


In this project we propose the development of an application that uses the most modern linguistic representations and natural language techniques to produce various resources that can be used by foreign language teaching systems. To put it to the test, we also designed a simple front-end system that allows us to consume these resources. In this way, we envision that our system, composed of the essential functionalities for the creation of learning resources, will be able to enter the market as an alternative to current applications for learning foreign languages.


In a world where annotated data is key for the development of research, products and services related to Artificial Intelligence, we propose an application to bring people together and gamify the monotonous task of Annotation. This tool intends to help bring teams toghether and help build extended teams where external people can contribute to the labelling of the data in a seamless fashion.


In this project we will develop a mobile application for the registration of Mexican indigenous languages that allows us to provide bilingual speakers (indigenous/Spanish), a simple and fast way to register translations, and make multimedia annotations in their language and indigenous variant. Our development will also provide options to take photos, upload representative images of the translated word, as well as the opportunity to record your own audios with the pronunciation of each word, and monitor the evolution of the alphabets. Thanks to this tool, field researchers will have more effective (and modern) working methods to interact with speaking communities, in addition to systematizing and sharing vocabularies with the rest of the scientific community.


Translatoria uses the power of synctactic trees to produce an open-source general purpose parser, which can translate human and machine readable structured texts into other human or machine languages the synctactic tree is fed with a rule-description-jsonschema during the first step then it can start translating texts right away. This project will help to automate translation of small-resource languages such as indigenous languages while other projects fetch data for machine-learning techniques.


As food-security and vaccine deliveries turn to be the most pressing issues of poverty and inequality during critical times, logistics become cornerstone for catalyzing social equality and inclusion of the people living in impoverished regions. Oftentimes, during times of health-crisis or war, the gap between the richest and the poorest grows wide. It is of paramount importance to have robust and powerful open-source algorithms that bring up the effectiveness and bring down the costs of the delivery of basic supplies to the poorest. These algorithms can be used by governments and NGOs to help alleviate the suffering of vulnerable populations.


A service that performs a validation of the written tweet, using Artificial Intelligence before publishing it. Indicating with 80% confidence whether the tweet has the correct wording to go viral. Significantly increases the reach and organic influence of each Tweet, improving your profile as an influencer and relying less on paid campaigns (effective in reach but not in influence).


A prototype that takes the dactylology of the Mexican sign language using sensors and tries to translate it to sound. This approach seeks to promote the inclusion of the deaf community in administrative environments where their rights could be exposed to risks. Currently, the prototype is merely mechanical. The integration of video signals is still pending.


Transdisciplinary Research for Augmented innovation – Technology is a commercial spin-off of TRAI-L. Devoted to de development of products and solutions for the industry, using transdisciplinary approaches and capabilities. TRAI-T also brings highly specialized human resources with an in-the-house BURST Methodology: in-and-out, high-value, low-footprint interventions to solve specific industry problems.

Early stage project. A social network dedicated to the specific needs of neighborhoods allowing for organization and inner commerce to emerge and continue in a sustained way.


Early stage project. A social network dedicated to the specific needs of neighborhoods allowing for organization and inner commerce to emerge and continue in a sustained way.

Supported projects.


    • A recent congress on the apllication of NLP and other computational techniques to the processing, preservation and revitalization of mesoamerican indigenous languages. In collaboration with Universidad Michoacana de San Nicolas Hidalgo (UMSNH).
  • Miyotl

    • It is a mobile application available on Google Play Store, AppGallery, App Store and web, suitable for all audiences, focused on education and dissemination of culture to rescue the 68 languages spoken in Mexico. It currently supports languages such as Chinanteco, Chocholteco, Chol, Mazatec, Maya, Mazahua, Mixe, Mixteco, Nahuatl, Otomí, Tzeltal, Tzotzil, Zapoteco, Totonaco, Zoque and soon the remaining 54 languages. Provides various texts or readings on cultural content such as gastronomy, clothing, stories, legends, dance, music, rituals, herbalism, nature, celebrations, “indigenous” literature, stories, uses and customs, poetry, among other genres that encompass national identity . It includes a dictionary for each language, with audiovisual content following the rules of the IPA (International Phonetic Alphabet). In collaboration with Universidad de Chapingo (UChap).
  • TzapincoBot

    • Early stage project that  works toward the construction of a robust and multifunctional agriculture robot.

    • App which helps to better understand and use rain patterns, helping to create strategies to prevent desertification of the hills in central Mexico.
  • DIAgnostika

    • Artificial Intelligence enhanced medical diagnosis. Based on the physician’s own notes it will raise alerts if it detects a possible illness diagnosis, helping wether to confirm the physicians supossitions or driving attention to unnoticed illness patterns.