Sharing ideas. Learning together. Building collaborations.
Research Spotlight is an informal series where teams from the Wellcome-funded cohort share what they're working on, what they've learned so far, and the questions and challenges emerging from their projects.
Whether you're in the early stages of your Wellcome-funded project or already generating results, this is a space to exchange perspectives, get constructive feedback, learn from one another's experiences, and discover opportunities for collaboration across the cohort.
Bring your research, your questions, and your curiosity.
Why join?
Share your progress: Introduce your project's objectives, approach, methods, and progress so far.
Get fresh perspectives: Exchange ideas and receive feedback from fellow Wellcome-funded teams.
Discuss challenges: Bring questions, uncertainties, and lessons learned from your ongoing work.
Find opportunities to collaborate: Discover shared interests, complementary expertise, and potential synergies across projects.
Learn across the cohort: Get inspired by the diverse approaches and experiences of teams working at the intersection of generative AI and mental health.
Join us to learn about Culturally Aligned Generative AI for Para-Counsellors, a project developing culturally aligned genAI to support frontline mental health care in Bangladesh. By combining lived-experience knowledge, culturally grounded causal models, and LLMs, the project aims to help para-counsellors better understand and respond to local expressions of distress. The team recently coined the term "participatory reasoning" in a newly accepted paper, which they will present alongside other findings.
Participatory Causal Reasoning: Toward Decolonizing Digital Mental Health in BangladeshLarge language models are increasingly used to support mental health care. However, local language or locally collected data does not necessarily change the assumptions through which they connect people’s circumstances, distress, and forms of support. This creates a deeper problem: who gets to shape the explanatory relationships that become computationally actionable? We examine Participatory Causal Reasoning (PCR) in community mental health care in Bangladesh through two connected phases. In Phase 1, people with lived experience, frontline practitioners, public health researchers, and clinicians reviewed and formalized these relationships into a Knowledge Graph. In Phase 2, we used this representation to ground AI-mediated counseling guidance. Stakeholder review changed what could responsibly become computational knowledge, separated interconnected problems, and exposed missing intervention and safety pathways. Across four models, grounded responses received better human evaluations even when reference-based similarity decreased. PCR shows how explanatory relationships can become a site of participation and governance.
Join the conversation to discover the project's approach and exchange ideas!
AURORA – Adaptive Understanding through Real-time Observation, Reporting, and Assistance
Presenter: Eric Storch (Baylor College of Medicine) and team
Join us to learn about AURORA, an innovative project exploring how AI and passive sensing can help us better understand and respond to changes in depression in real time. AURORA combines conversational genAI with multimodal sensing to capture both changes in depressive symptoms and the lived experiences behind them. The goal is to move beyond traditional, periodic assessments toward more responsive, personalized, and equitable mental health care.
Join the conversation to discover the project’s approach, exchange ideas, ask questions, and explore potential connections with your own work.
Date: 17 December 2026
Time: 14:00–15:00 UTC · 8:00 AM–9:00 AM US Central time (CST)
ARIADNE: ARtificial Intelligence-based Assessment to Detect iNdividuals with Emerging psychosis risk
Presenter: Dominic Oliver (University of Oxford) and team
In this session, join us to learn about ARIADNE, an innovative project developing a voice-based interview system powered by generative AI to improve the identification of young people at clinical high risk for psychosis (CHR-P). By enabling natural conversations, asking relevant clinical questions, and generating tailored reports for patients, clinicians, and researchers, ARIADNE aims to make assessments more accessible, efficient, and empowering, while supporting early intervention and preventive care.
Join the conversation to discover the project's approach and exchange ideas!
Date: 19 January 2027
Time: 14:00–15:00 UTC · 3:00 PM–4:00 PM UK time (GMT)
FarajaMH: Culturally Adapted Generative AI Model for Mental Health Screening in Kenya and Tanzania
Presenter: Bylhah Mugotitsa and Lucy Akinyi
Join us to learn about FarajaMH, a project developing a culturally grounded, bilingual generative AI model to improve the early screening of depression, anxiety, and psychosis in Kenya and Tanzania. By leveraging mental health data, clinical notes, and conversational data in English and Swahili, FarajaMH aims to better capture culturally specific expressions of psychological distress and make mental health screening more accessible, accurate, and equitable. The team will present “A Culturally Grounded Framework for Mental Health Symptom Detection,” sharing their approach and vision for AI-assisted mental healthcare in East Africa.
Join the conversation to exchange ideas and ask questions.
Each session lasts one hour and is designed to leave plenty of room for discussion.
5 minutes: Welcome and introductions
20–30 minutes: Research presentation or project walkthrough
20–30 minutes: Q&A, feedback, and group discussion
5 minutes: Key takeaways and wrap-up
Presenters can share their project's objectives, research approach, progress to date, preliminary findings, lessons learned, or challenges they'd like input on. The format is intentionally informal and interactive: the goal is to learn from one another's Wellcome-funded work, rather than deliver a formal seminar. The format is flexible. Presenters can share slides, demonstrate a tool, discuss preliminary findings, or bring questions to explore with the group.
Who can join?
All members of the Wellcome-funded cohort are welcome, regardless of their research area or stage of project development.
You don't need to prepare anything in advance. Come ready to listen, share your perspective, ask questions, and learn from the experiences of other teams.
Present your project
We'd love to hear from teams across the cohort! Research Spotlight is an opportunity to share and discuss your ongoing Wellcome-funded project, including:
Your project's objectives, research questions, and approach
Progress to date, including preliminary findings or emerging insights
Methods, tools, or technical approaches you're developing or using
Challenges, uncertainties, or lessons learned along the way
Questions where feedback from other cohort teams could be helpful
Opportunities for synergies, knowledge exchange, or collaboration with other funded projects
Interested in presenting? Contact the organizers with a few dates that work for your team.
Past sessions & resources
After each session, we'll share key takeaways, presentation materials where available, useful links, and potential collaboration opportunities.
Let's learn from one another and strengthen connections across the cohort.
Add a caption...
Session 01 · 20 October, 2026
Bridging the Treatment Gap with AI-Enhanced Supervision (BRIDGE-AI)
Presenters: Julia R. Pozuelo (Harvard), Anuja Lahiri (Sangath), and Ryan Christopher Louie (Stanford)
Learn about BRIDGE-AI, an initiative using AI to support counselors and improve access to evidence-based psychological treatments in India.