Multilingual Natural Language Processing & Sentiment Analysis

The contract involves developing and applying advanced AI and machine learning models to process and analyze text in Spanish and indigenous Guatemalan languages, including K’iche’ and Kaqchikel, with the goal of detecting sentiment, identifying thematic patterns, and uncovering emerging social narratives. This work is critical for understanding public opinion and cultural discourse in Guatemala, particularly among communities where indigenous languages are spoken, ensuring that digital analysis tools are inclusive and linguistically accurate. The project requires robust linguistic engineering, culturally informed training data, and models capable of handling linguistic diversity, dialectal variation, and low-resource language conditions. The effort is subcontracted under the U.S. Embassy in Guatemala City, part of the Department of State, with performance based in Guatemala City, ZIP 01016. The opportunity was posted on July 16, 2026, and Responses are due by August 25, 2026. It falls under the NAICS code 541715, which classifies it as research and development in the physical, engineering, and life sciences, emphasizing computational linguistics and natural language processing. The contract seeks specialized expertise in multilingual NLP systems that can operate effectively in under-resourced language environments, with a strong emphasis on ethical AI practices and local language preservation.