Our Projects

From idea to implementation, our projects translate research into tools that support better decision-making.”

African Language Technology

Bridging Language Gaps in the Parish Dev't Model

Signed a 5-year MoU with MoICT&NG to integrate translation tools into gov’t citizen infrastructure.

African Language Technology

Real-Time Urban Noise Intelligence in East Africa

We deployed custom, Edge AI-powered sensors across Kampala and Entebbe to classify urban soundscapes.

Environmental Monitoring

Voice AI for Agriculture with Amara Hub

In partnership with Amara Hub, we developed a voice-first agricultural information system.

African Language Technology

Amplifying Citizen Voices with TRAC FM

Developed an automated speech-to-text & translation pipeline for TRAC FM’s radio call data.

African Language Technology

Democratising Citizen Feedback with SEMA

Sunbird AI integrated our speech-to-text and translation pipeline directly into SEMA’s devices.

Data-Driven Governance

DataCities — Data-Driven Urban Governance

We are helping Fort Portal & Jinja assess & build their data ecosystems for data-driven urban policy.

Data-Driven Governance

AI for Maternal & Reproductive Health (HASH)

Co-founded HASH a research hub uniting medical professionals, computer scientists, & policy experts.

Data-Driven Governance

Gender and Responsible AI Network (GRAIN)

Co-founded GRAIN, with IPAR (Senegal) & CSEA (Nigeria) to advance gender equality through locally grounded AI. 

Data-Driven Governance

Social Listening for COVID-19 Pandemic Response

Worked with Uganda’s national COVID-19 Task Force & KCCA to analyse public reaction to gov’t health measures.

Multilingual Digital Services with Cente-Tech

Integrated machine translation into its messaging platform, enabling multilingual SMS & USSD services.

African Language Technology

AI-Powered Call Centre Intelligence with dfcu Bank

Implemented a speech recognition and translation pipeline to transcribe and translate dfcu’s call centre recordings.

African Language Technology

AI for Inclusive Education with Back-Up Uganda

Sunbird AI’s speech-to-text technology in both English and Acholi is embedded into Back-Up Uganda’s workflows.

African Language Technology

African Language Technology

PDPO Complaint Streamlining

Sunbird AI collaborated with PDPO to build an AI-powered complaint classification system.

African Language Technology

Bridging Language Gaps in the Parish Dev't Model

Signed a 5-year MoU with MoICT&NG to integrate translation tools into gov’t citizen infrastructure.

African Language Technology

Real-Time Urban Noise Intelligence in East Africa

We deployed custom, Edge AI-powered sensors across Kampala and Entebbe to classify urban soundscapes.

Environmental Monitoring

Voice AI for Agriculture with Amara Hub

In partnership with Amara Hub, we developed a voice-first agricultural information system.

African Language Technology

Amplifying Citizen Voices with TRAC FM

Developed an automated speech-to-text & translation pipeline for TRAC FM’s radio call data.

African Language Technology

Democratising Citizen Feedback with SEMA

Sunbird AI integrated our speech-to-text and translation pipeline directly into SEMA’s devices.

Data-Driven Governance

DataCities — Data-Driven Urban Governance

We are helping Fort Portal & Jinja assess & build their data ecosystems for data-driven urban policy.

Data-Driven Governance

AI for Maternal & Reproductive Health (HASH)

Co-founded HASH a research hub uniting medical professionals, computer scientists, & policy experts.

Data-Driven Governance

Gender and Responsible AI Network (GRAIN)

Co-founded GRAIN, with IPAR (Senegal) & CSEA (Nigeria) to advance gender equality through locally grounded AI. 

Data-Driven Governance

Social Listening for COVID-19 Pandemic Response

Worked with Uganda’s national COVID-19 Task Force & KCCA to analyse public reaction to gov’t health measures.

Multilingual Digital Services with Cente-Tech

Integrated machine translation into its messaging platform, enabling multilingual SMS & USSD services.

African Language Technology

AI-Powered Call Centre Intelligence with dfcu Bank

Implemented a speech recognition and translation pipeline to transcribe and translate dfcu’s call centre recordings.

African Language Technology

AI for Inclusive Education with Back-Up Uganda

Sunbird AI’s speech-to-text technology in both English and Acholi is embedded into Back-Up Uganda’s workflows.

African Language Technology

African Language Technology

PDPO Complaint Streamlining

Sunbird AI collaborated with PDPO to build an AI-powered complaint classification system.

Scroll to Top
Bridging Language Gaps in the Parish Development Model (MOICT)

Uganda’s Parish Development Model (PDM) targets citizens in subsistence economies, most of whom don’t speak English.  The PDM aims to uplift rural communities by providing economic opportunities. However, language barriers, with many beneficiaries unable to speak English, hinder effective communication.

Sunbird AI partnered with the Ministry of ICT and National Guidance in Uganda to support the Parish Development Model (PDM) initiative. To address this challenge, Sunbird AI is providing advanced machine translation technology. By integrating our tools into the PDM information system, we facilitate communication between government officials and citizens. This ensures that information is accessible to all, and that feedback can be effectively gathered and understood.

Real-Time Urban Noise Intelligence in East Africa: Deploying Africa’s first real-time noise monitoring network and building the world's largest urban acoustic dataset.

Africa is rapidly urbanising, and this comes with new public health hazards. Some problems are easily overlooked because of a lack of data. Noise pollution, for example, may seem merely an annoyance, but in fact is a major health concern. It is linked to deaths from ischaemic heart disease (as the body’s nervous system is constantly ‘on alert’ in noisy places), causes learning impairment in children who cannot sleep well, and hinders the recovery of patients recovering from illness who need to rest. In Europe, noise pollution is linked to one million healthy years of life lost each year. Across Africa, city planners have historically lacked the data needed to measure and manage acoustic pollution. You cannot fix what you cannot measure.

In partnership with KCCA and  Entebbe City Authority, Sunbird AI built the solution from the ground up. We deployed custom, Edge AI-powered sensors across Kampala and Entebbe to classify urban soundscapes into 19 categories 

This initiative has yielded an urban noise dataset, making it the largest urban acoustic dataset in existence that we know of. Peer-reviewed in Nature’s Scientific Data, our open-access dataset and live dashboard (noise.sunbird.ai) give policymakers, researchers, and regulators the evidence needed to design quieter, healthier, and more livable cities.

Voice AI for Agriculture with Amara Hub

In northern Uganda, smallholder farmers need agricultural advice and market information in their local languages. However, extension workers are scarce, internet connectivity is low, and text-based tools require literacy skills many farmers do not have.

To bridge this gap, Sunbird AI partnered with Amara Hub, a non-profit empowering underserved communities in northern Uganda, to build a voice-first agricultural information system. Powered by Sunbird Speech and Translation APIs, the platform allows farmers to receive and share vital updates simply by speaking in their native languages.

This inclusive AI connects rural farmers to essential data without requiring English literacy, offering a practical model for voice-based community tools that can be replicated across Uganda’s agricultural sector

Amplifying Citizen Voices with TRAC FM

23% of Uganda’s population is illiterate, and most lack internet access. Radio is how millions engage in public debate. TRAC FM receives 20,000 responses weekly across Ugandan languages. Manual transcription and translation at this scale are tedious and expensive.

Sunbird AI developed an automated speech-to-text and translation pipeline for TRAC FM’s radio call data. Spoken responses are transcribed in the caller’s language, translated into English, and processed for sentiment and topic, converting raw voice data into structured insight for journalists, civil society, and government.

Citizens with no literacy or internet access can now contribute meaningfully to public policy debates. Decision-makers receive structured, data-driven feedback at scale.

Democratising Citizen Feedback with SEMA

Citizens using SEMA’s public feedback devices in hospitals, schools, and government offices could only give star ratings no reason, no voice, no language choice; these devices only captured “happy” or “unhappy” emojis. Feedback was thin and limited to those literate in English.

Sunbird AI integrated our speech-to-text and translation pipeline directly into SEMA’s devices. Citizens can now record voice feedback in their preferred Ugandan language and articulate their issues: for example, to report that the drugs they needed were out of stock.

The audio is automatically transcribed, translated into English, and analysed for sentiment, delivering rich, actionable insight to decision-makers.

The impact is broader participation across language and literacy lines. AI-powered sentiment analysis gives institutions an accurate understanding of citizen experience. A new standard for inclusive public feedback in Uganda.

DataCities

Uganda’s transition of 15 municipalities into regional cities brings major economic opportunities, but it also creates rapid population growth, infrastructure pressure, and rising service demands. Many of these local governments operate without the data systems or analytical tools needed for effective urban planning.

To address this gap, Sunbird AI joined the DataCities initiative alongside UN Global Pulse and ToroDev. The project works directly with city authorities to establish practical, resilient data ecosystems and enabling environments for data-driven urban policy. 

 

Link to DataCities website

Hub for Artificial Intelligence in Maternal, Sexual and Reproductive Health(HASH)

Scaling Responsible AI in African Maternal & Sexual Health

Sub-Saharan Africa bears a heavy share of the global burden in maternal mortality, adolescent reproductive health risks, and untreated sexually transmitted infections. While digital health data is expanding across the region, local health practitioners and researchers rarely have access to specialized machine learning expertise. Consequently, life-saving AI innovations struggle to move from concepts into active clinical use.

To solve this, Sunbird AI co-founded the Hub for Artificial Intelligence in Maternal, Sexual, and Reproductive Health (HASH) alongside the Infectious Diseases Institute (IDI) and Makerere University AI Lab. Funded by Canada’s International Development Research Centre (IDRC) and Sida through the AI4D Africa program, HASH is a Pan-African initiative connecting medical professionals, computer scientists, and policy experts across Anglophone and Francophone Africa.

As the core technical partner, Sunbird AI builds and maintains the hub’s engineering foundation. Sunbird AI provides machine learning architecture, establishes strict data privacy frameworks, and guides local data integration.

https://hash.theacademy.co.ug/

Gender and Responsible AI Network (GRAIN)

Sunbird AI co-founded GRAIN, the Gender and Responsible Artificial Intelligence Network with IPAR (Senegal) and CSEA (Nigeria). The network advances gender equality and inclusion in sub-Saharan Africa through responsible, locally grounded AI. GRAIN’s three pillars: building an inclusive research network; capacity building and advocacy on responsible AI for gender equity; and implementing a research agenda for gender-responsive AI in Africa.

 grain-africa.org

Social Listening for COVID-19 Pandemic Response

At the start of the COVID-19 pandemic in 2020, Sunbird AI worked with Uganda’s national COVID-19 Task Force and KCCA to analyse public reaction to government health measures. Using NLP on anonymised Ugandan Facebook and Twitter data, we built a social media dashboard tracking public sentiment toward lockdowns, school closures, and SOPs. Findings were delivered directly to Ministry of Health partners to inform the national response.

Multilingual Digital Services with Cente-Tech

Cente-Tech, a Ugandan ICT solutions provider, wanted to expand its SMS and USSD digital messaging platform to serve communities across Uganda, delivering financial literacy content and educational initiatives. A single-language platform was excluding the majority of potential beneficiaries who don’t communicate primarily in English.

Cente-Tech integrated Sunbird AI’s machine translation model directly into its messaging platform, enabling SMS and USSD services to reach users in multiple Ugandan languages. Citizens can now receive and respond to financial literacy and educational content in their preferred language

AI-Powered Call Centre Intelligence with dfcu Bank

DFCU  Bank, one of Uganda’s leading financial institutions, receives thousands of customer calls across multiple languages. Manual review of call centre data at this scale was impossible, meaning valuable customer feedback and recurring issues were going undetected and unaddressed.

Sunbird AI implemented a speech recognition and translation pipeline to transcribe and translate dfcu’s call centre recordings across Ugandan languages and English. The system converts spoken customer interactions into structured, searchable text, enabling dfcu to analyse feedback patterns, identify common queries, and detect service gaps at scale.

dfcugroup.com

AI for Inclusive Education with Back-Up Uganda

Back-Up Uganda supports students with learning difficulties across Uganda, including in Acholi-speaking communities in the north. Creating accessible educational content and communicating effectively with students, parents, and community members required tools that worked in local languages tools that didn’t exist.

Sunbird AI’s speech-to-text technology in both English and Acholi is embedded into Back-Up Uganda’s workflows. The organisation uses it to create accessible educational content, support parents and teachers with students with reading and language challenges, and raise awareness about learning disabilities within Acholi-speaking communities.

Parents with children with learning difficulties in northern Uganda can now access educational support in their language. A pioneering use of Acholi speech technology for inclusive education and a demonstration that local-language AI can reach the most underserved learners.

 backupuganda.org



Streamlining Data Protection Complaints with PDPO

Uganda’s Personal Data Protection Office (PDPO), established under the Data Protection and Privacy Act of 2019, was facing a growing backlog of complaints from citizens. Manual classification of complaints was slow and inconsistent, delaying resolution and undermining public trust.

Sunbird AI collaborated with PDPO to build an AI-powered complaint classification system. We annotated complaint data to train a machine learning model that automatically categorises incoming complaints by type and severity. Where standard models struggled with class imbalance across complaint categories, we incorporated large language models (LLMs) for more nuanced classification, significantly improving accuracy.

  pdpo.go.ug