Amsterdam – CommuniCity

Confidential real-time translation for Safe Home conversations

Confidential real-time translation for Safe Home conversations

Challenge: Amsterdam: How to support residents facing domestic violence?
Target Group: Families

Veilig Thuis (Safe Home) Netherlands receives around 125,000 reports of domestic violence and child abuse each year. In many cases, language barriers make it hard to offer the right help, especially when no interpreter is available. Generic translation tools often fall short or raise privacy concerns.

To address this, an AI interpreter is being developed to support Safe Home staff during home visits and phone calls. The app translates speech in real time and shows written transcriptions, all with privacy in mind.

Co-creation was a key part of the pilot. Contextual words like legal and social terms were collected online and during a workshop with four staff members. Two roleplaying sessions were held at the Amsterdam office, each with two groups of staff simulating real-life cases. Turkish was tested in the first session; Spanish was added later.

A privacy workshop included eight participants, including compliance officers, to make sure the tool aligns with regulations. Feedback from all sessions helped improve the interface, translations, and overall usability.

Next steps include further development and testing, aiming to support Safe Home organizations across the Netherlands and eventually Europe, helping more people receive the care they need, regardless of language.

Access to Healthcare Chatbot (A2HC Chatbot)

Access to Healthcare Chatbot (A2HC Chatbot)

Challenge: Amsterdam: Wildcard – How can an AI solution solve a challenge faced by a marginalised group?
Target Group: Migrants

The pilot focused on co-developing a healthcare chatbot to support undocumented (economic) migrants, refugees, and rejected asylum seekers in Amsterdam. The objective was to improve access to reliable healthcare information by creating a tool that integrates institutional knowledge with the lived experiences of the foreign community.

Development involved a participatory process with four co-creation sessions: two with NGOs such as Wereldhuis, and two with undocumented community members. The chatbot is based on real-life experiences with accessing the healthcare system. Around 13 individuals contributed to the sessions, representing the undocumented economic migrants, refugees, and asylum seekers community.

From a technical perspective, the chatbot employs a full-stack development approach and utilises AI models from OpenAI for both text and voice processing. Strict privacy protocols are maintained, ensuring no sensitive data is stored and all community input is anonymised.

Co-creation with participants from marginalised groups fostered a sense of ownership and trust. To facilitate this, safe spaces were provided along with food and logistical support, ensuring respectful and positive engagement for both NGOs and community members.

The pilot concluded with a chatbot that is technically ready and contextually relevant, with plans to continue development and community engagement beyond the project’s duration. Future collaboration is envisaged with Dutch and European partners to scale or adapt the solution for other cities and contexts.

Prevent misunderstanding with a context-specific AI interpreter

Challenge: Amsterdam: Wildcard – How can an AI solution solve a challenge faced by a marginalised group?
Target Group: Migrants

To support the Municipality of Amsterdam in helping over 8,000 statusholders each year, Mokin.nl is building an AI interpreter: a real-time speech-to-speech translation app designed for public service conversations.

The goal is to reduce misunderstandings between municipal staff and statusholders, especially when no translator is available. Unlike generic tools, the app uses context-specific words like “bijstandsuitkering” or “belastingdienst,” which are hard to translate and often unknown to newcomers.

Co-creation is a key part of the project. Together with statusholders, the team collects important Dutch words and concepts to train the AI. A roleplaying workshop was held with a statusholder and municipal staff to test real-life situations. Feedback from this session helped improve the app’s language use and features.

The pilot shows better understanding in conversations and more confidence from both sides. The updated word list, new language options, and privacy-aware features all contribute to smoother communication.

Next steps include further testing in Amsterdam and preparing to scale the tool across Europe, where it can support integration for over one million refugees every year.

EmpathyBot: Guiding Parents in High-Conflict Divorces

EmpathyBot: Guiding Parents in High-Conflict Divorces

Challenge: Amsterdam: How can a chatbot support parents navigating a high-conflict divorce?
Target Group: Families

The Empathy Bot pilot aimed to develop an AI-powered chatbot to support professionals working with families experiencing high-conflict divorces, with a focus on the emotional well-being of children. The objective was to create a tool that could generate empathetic, context-aware narratives and guidance, enhancing how caretakers assist parents during sensitive situations.

Co-creation was central to the process. Collaboration with caretakers from Levvel shaped the chatbot’s tone, ethical boundaries, and core functionality. Initial questionnaires gathered insights from a broader sample, while weekly feedback sessions with approximately 10 professionals guided iterative development. Key features—such as duo sessions for co-parents and personalized storytelling—emerged directly from this collaborative process.

The result is a functional prototype, tested in a controlled environment, which reached Technology Readiness Level 6. While not yet ready for direct use by families due to privacy considerations, the solution has shown strong potential as a professional support tool.

Next steps include enhancing privacy safeguards, refining the user interface, and expanding the interactive and emotional capabilities of the platform. A continued partnership with Levvel is planned, with the goal of long-term development and formal co-ownership of the tool.

Roadmap Canvas for the Regenboog Groep

Roadmap Canvas for the Regenboog Groep

Challenge: Amsterdam / Regenbooggroep: How to support the “joint journey” of the volunteers and participants of the Rainbow Group?
Target Group: individuals who are homeless, in debt, addicted or otherwise socially isolated and volunteers working with them

The Maatjeskaarten pilot aimed to support buddy pairs of The Rainbow Group (Regenbooggroep) with an easy-to-use tool that fosters meaningful conversations and shared activities. Starting from co-creation workshops, featuring 6-8 participants from the Rainbow Group, the team shifted from a digital roadmap idea to printed cards with conversation starters, activity ideas, and light challenges.

During the pilot, coordinators distributed the cards and helped select relevant sets. A digital version was also available but less used. Feedback highlighted the cards’ flexibility and positive focus, though limited testing time and varied engagement posed challenges.

Next steps involve collecting more feedback, refining the cards, and considering wider rollout within the organisation.

Culturally Sensitive Tool for Moroccan Migrants with Dementia

Picture of 4 people by Henrik Terävä. Pilot: Henrik - How can migrants with dementia stay fit and learn to train their memory?

Culturally Sensitive Tool for Moroccan Migrants with Dementia

Challenge: Amsterdam: How can migrants with dementia stay fit and learn to train their memory?
Target Group: Migrants with dementia

This pilot project was carried out in Amsterdam using PALL0, an interactive, culturally agnostic ball designed to support individuals with dementia. The initiative involved collaboration with Ouderen Kliniek and Amsterdam UMC. The objective was to explore how physical and cognitive activity could be promoted through simple, intuitive play.

Participants included first-generation Moroccan migrants in the early stages of dementia. Sessions were facilitated by caregivers and health professionals in familiar care environments. Co-creation activities included observational feedback from staff and users, informal interviews, and iterative testing to refine the use of PALL0 in real-life settings. Approximately 10 people were involved, with half being individuals in the early stages of dementia. PALL0 offered multisensory engagement through light, sound, and movement, encouraging interaction without relying on screens or complex instructions.

The pilot demonstrated that low-barrier, tactile technology can contribute meaningfully to dementia care by fitting naturally into daily routines.

The findings suggest further exploration of PALL0 in long-term care and rehabilitation settings could be beneficial for maintaining cognitive and emotional engagement among aging populations. Next steps include broader piloting.

Fonetic@ for Amsterdam Central Station: Safe AI translation

Challenge: Amsterdam: Wildcard – How can an AI solution solve a challenge faced by a marginalised group?
Target Group: Persons with disabilities

For many individuals, navigating public spaces is an effortless activity that requires little conscious attention. It is often only when encountering a different culture or when sensory disruptions occur that it becomes apparent that public spaces are most accessible when designed with diverse needs in mind.

In this pilot, Amsterdam’s Central Station was enhanced with a ‘hyper-local AI’ system. This AI enables members of the local community to send and receive information through their smartphones, using their preferred language. The solution was co-created with a hearing-impaired participant to ensure accessibility considerations were embedded from the outset. Testing took place within a restaurant at the station, an environment presenting acoustic challenges representative of the wider station context.

Girls Meet Up

Challenge: Amsterdam: Wildcard – How can an AI solution solve a challenge faced by a marginalised group?
Target Group: Children and youth

The Girls Meet Up/MeidenApp pilot aimed to increase sports participation among girls aged 12–14 in Amsterdam-Zuidoost by combining digital engagement with community-based support. The key objective was to understand what motivates girls to stay active and create a playful, accessible way for them to engage with sports, both in clubs and at home.

The pilot involved four co-creation sessions with 30 participants, including local coaches and youth workers (from ZuidOost United and community organizations), focusing on how girls interact with training, what motivates them, and how to simplify the coaches’ tasks. Insights from these sessions led to the development of a web-based platform where girls could join local sports events and complete home-based “exercises” — short challenges with deadlines, descriptions, and sometimes videos.

Three structured training sessions were created, each linked to a preparatory exercise. To join a training, girls were first asked to complete the exercise by watching a short video and reading the description. This approach helped build motivation and readiness. Coaches also gained visibility into how girls completed the challenges and how quickly they responded, making it easier to track engagement and adjust support as needed.

Next steps include adding a points-based reward system and expanding the pilot with local partners.

Lighting System for Neurodivergent Populations

Lighting System for Neurodivergent Populations

Challenge: Amsterdam: Wildcard – How can an AI solution solve a challenge faced by a marginalised group?
Target Group: Neurodivergent

Neurodivergent individuals are often highly sensitive to environmental stimuli such as light, sound, and smell. This pilot project explored how lighting conditions could be adapted in high-focus professional environments to better support neurodivergent users. Conducted in collaboration with two audio-visual (AV) surveillance units of the Dutch Police (Rotterdam), the project set out to investigate how overhead lighting might be optimized to reduce sensory stress and improve comfort and performance.

In Stage 1, researchers gathered input from participants about their lighting preferences. The neurodivergent community had 12 representatives present during the co-creation process. The feedback highlighted diverse needs for light intensity and temperature based on personal sensitivities and work conditions. Building on these insights, Stage 2 of the pilot involved developing a software-based solution that allows each individual to personalize their lighting experience. The system incorporates both manual control and AI-driven adjustments that account for occupancy levels and ambient light conditions, striking a balance between individual comfort and operational efficiency.

T-APPS B.V. – Skendy: an assistant for your documents

T-APPS B.V. – Skendy: an assistant for your documents

Cross-border Pilot with Amsterdam

T-APPS B.V. – Skendy: an assistant for your documents

The CommuniCity project Skendy piloted an initiative designed to improve support mechanisms for expats and refugees through an advanced mobile app. This pilot entailed over 40 hours of co-creation sessions with expats, refugees, and various stakeholders such as IAMSTERDAM, students, and local communities in Amsterdam and Prague. The objective was to identify and address the unique challenges faced by these groups through practical, user-focused solutions.

During the piloting process, an event in Prague played a pivotal role in testing and refining these solutions based on participant feedback. The initiative’s core developments included the design and implementation of a new user interface and the integration of an AI-powered chatbot. This chatbot, intended as the central feature of the app, provides personalized assistance and access to a comprehensive knowledge base tailored for both Amsterdam and Prague.

Additionally, the pilot phase included the creation of a demo version of the app specifically for refugees. This demo, developed in collaboration with refugee representatives, aimed to ensure the platform effectively meets their specific needs, although it remains in the testing stages. The pilot yielded promising results: engaging target communities, developing a user-friendly digital interface, and implementing a functional AI assistant that enhances the app’s utility. It also led to the acquisition of new talent to facilitate faster growth and increased support capabilities.

The next steps following the pilot include leveraging the insights gained to refine the app further. These efforts highlight the project’s dedication to continuous enhancement and community involvement in developing services that directly cater to the needs of expats and other end-users.