Boda AI
Factory

Research background

The research experience our founders bring.

Our founders' work in applied AI spans their research affiliations, including the Makerere Artificial Intelligence Lab, Sunbird AI, the University of Washington, Mozilla Common Voice, UN Global Pulse, and Oracle Research. A sample of that work is below, credited to the person who led or contributed to it.

  • Healthcare

    Ocular

    A smartphone-based diagnostic microscopy tool for detecting malaria, tuberculosis, and cervical cancer.

    Solomon Nsumba
  • Speech & text

    SALT-31

    An open machine translation benchmark covering 31 Ugandan languages.

    Solomon Nsumba
  • Speech & text

    Sunflower

    Multilingual language models built for Ugandan languages.

    Solomon Nsumba
  • Infrastructure

    Urban-Noise-Uganda-61K

    An open acoustic dataset of urban sound recorded in Kampala and Entebbe.

    Solomon Nsumba
  • Healthcare

    Wearable camera for medication-error detection

    An AI-enabled wearable camera system for detecting medication errors, developed with researchers at the University of Washington.

    Solomon Nsumba
  • Agriculture

    Crop & livestock disease-vector surveillance

    Large-scale datasets supporting surveillance of crop and livestock disease vectors across African contexts.

    Mutembesa Daniel
  • Agriculture

    Mobile sensor deployment optimization

    Multi-objective optimization strategies for deploying mobile sensors in resource-constrained environments.

    Mutembesa Daniel
  • Finance

    Social-knowledge-informed credit scoring

    Credit scoring for SMEs that draws on social data, extended into algorithmic fairness for equitable digital micro-credit and micro-insurance design.

    Mutembesa Daniel
  • Speech & text

    Speech recognition for SDG monitoring

    Speech recognition models supporting social listening for Sustainable Development Goal monitoring, built with UN Global Pulse.

    Jonathan Mukiibi
  • Speech & text

    Gender bias in automatic speech recognition

    Research identifying and addressing gender bias in automatic speech recognition model performance.

    Jonathan Mukiibi
  • Speech & text

    Inclusive open-source speech datasets

    Privacy-conscious, open-source speech datasets for low-resourced languages, built with the Mozilla Common Voice project.

    Jonathan Mukiibi
  • Speech & text

    Morphological complexity in children's speech

    Research on multilingual transfer learning for inclusive speech tools in children's education, with Oracle Research.

    Jonathan Mukiibi

Working on a related problem?

Collaborate with Boda AI