Belén Saldías
Also known as Belen Saldias, Belén Saldías Fuentes, Belén Saldías-Fuentes
Belén Saldías is an assistant teaching professor at the University of Washington Information School whose research focuses on responsible machine learning and human-AI collaboration. She earned engineering and machine-learning degrees in Chile, led data-science work at Falabella, and joined MIT’s LSM/CCC group in 2018. Her work has ranged from Tweet Moodifier and Gobo to machine-translation evaluation, qualitative analysis, and child-centered content moderation. With Irti Haq, she studied how dialect and stated user profiles affect language-model responses in 2026; by May 2026, she had been selected as an inaugural recipient of the Google Higher Ed Faculty AI Fellowship for North America.
- Now at
- University of Washington
Timeline
-
2012–2017
Engineering and machine learning at PUC
From 2012 to 2017, she studies engineering and computer science at PUC and earns a B.S. She completes an M.Sc. in machine learning at PUC in 2017.
-
2016–2018
Harvard research and Falabella data science
In 2016–2017, she is a research assistant at Harvard’s IACS. From 2016 to 2018, she works as a lead data scientist at Falabella.com. She is also an adjunct professor at the Pontifical Catholic University of Chile from 2017 to 2018.
-
2018–2020
Tweet Moodifier, Gobo, and personal narratives
She joins MIT’s LSM/CCC group as a research assistant in 2018. In 2019, she coauthors Tweet Moodifier, a Chrome extension that marks emotional content in Twitter feeds, and Gobo, an open-source social-media browser for filtering content across platforms. With Deb Roy, she studies spoken personal narratives in 2020 using a corpus of 10,296 clauses from 594 video transcripts.
-
2021–2022
Google Translate and machine translation evaluation
In 2021, she is a summer research intern with Google Translate in Montreal. In 2022, she coauthors Toward More Effective Human Evaluation for Machine Translation, which uses stratified sampling and control variates to improve estimates from a fixed annotation budget.
-
January 2023
MIT–Chile workshops and a graduate award
In January 2023, she leads MIT–Chile workshops on human-centered machine learning, natural language processing, data visualization, and causal inference. She receives an MIT Graduate Women of Excellence Award in 2023.
-
2024–2025
SenseMate and child-centered moderation
In 2024, she coauthors SenseMate, a human-AI platform for qualitative data analysis. Her 2024 paper on child-centered content exposure and moderation sets out design considerations for a family-guided system. She completes her Ph.D. at MIT’s Media Lab and the Center for Constructive Communication in 2025.
-
2025
Assistant teaching professor at UW
She joins the University of Washington Information School as an assistant teaching professor for the start of the 2025–26 academic year.
-
February 2026
Affiliated faculty at UW
In February 2026, the Center for an Informed Public names her an affiliated faculty member for a two-year renewable term. Her research focuses on human-centered machine learning, human-AI collaboration, user agency, and governance.
-
April 2026–May 2026
Dialect vs Demographics and a fellowship
With Irti Haq, she submits Dialect vs Demographics in April 2026, analyzing over 24,000 responses from two open-weight LLMs to compare explicit user profiles with dialect cues. By May 2026, she is selected as an inaugural recipient of the Google Higher Ed Faculty AI Fellowship for North America.
-
As of October 2026
Assistant teaching professor at UW
She is an assistant teaching professor at the University of Washington Information School and teaches Generative AI Ethics and Data Justice. On October 2, 2026, she participates in the Society + Technology at UW symposium.
Career
- Since 2026University of WashingtonAffiliated Faculty Member, Center for an Informed Public
- Since 2025University of WashingtonAssistant Teaching Professor
- 2021GoogleSummer Research Intern, Google Translate
- From 2018Massachusetts Institute of TechnologyResearch Assistant, LSM/CCC group
- 2017–2018Pontifical Catholic University of ChileAdjunct Professor
- 2016–2017Harvard UniversityResearch Assistant, IACS
- 2016–2018Falabella.comLead Data Scientist
Education
- 2025Massachusetts Institute of TechnologyPh.D., Media Lab and Center for Constructive Communication
- 2017Pontifical Catholic University of ChileB.S. in Engineering and Computer Science
- 2017Pontifical Catholic University of ChileM.Sc. in Machine Learning
Sources and credits
Photo: We have no published photo of Belén Saldías yet. Send us a free-licensed photo.
- Belén Saldías
- Belén Saldías - MIT Center for Constructive Communication
- Tweet Moodifier: Towards giving emotional awareness to Twitter users
- Gobo: A System for Exploring User Control of Invisible Algorithms in Social Media
- Exploring aspects of similarity between spoken personal narratives by disentangling them into narrative clause types
- Toward More Effective Human Evaluation for Machine Translation
- MIT–Chile Research Workshops | Human-Centered Artificial Intelligence and Data Visualization
- Belén Saldías Fuentes and Ufuoma Ovienmhada honored during 2023 MIT Graduate Women of Excellence Ceremony
- SenseMate: An Accessible and Beginner-Friendly Human-AI Platform for Qualitative Data Analysis
- Designing Child-Centered Content Exposure and Moderation
- iSchool Directory | Information School | University of Washington
- Four scholars to join the Information School faculty | Information School | University of Washington
- CIP announces appointments for 6 new faculty members | Center for an Informed Public
- Dialect vs Demographics: Quantifying LLM Bias from Implicit Linguistic Signals vs. Explicit User Profiles
- May 19, 2026 | Information School | University of Washington
- Society + Technology event assembles array of UW scholars | Information School | University of Washington
