Dom Sweeting
I’m a psychology researcher building a trauma‑informed chatbot prototype to learn what helps (and what doesn’t) in real‑world conditions. My aim is to create a tool that is usable, ethically cautious, and culturally respectful.
A calm, bounded conversation tool designed for early support — especially in war‑affected contexts. Plain language. Stop anytime.Спокійний інструмент для розмови з чіткими межами — для ранньої підтримки, зокрема в умовах війни. Проста мова. Можна зупинитися будь-коли.
Pilot research prototype. Short sessions. Plain language. You can stop at any time.Пілотний дослідницький прототип. Короткі сесії. Проста мова. Можна зупинитися будь-коли.
A brief, honest introduction — and what I’m trying to build.Коротко й чесно — хто я, і на що я не претендую.
I’m a psychology researcher building a trauma‑informed chatbot prototype to learn what helps (and what doesn’t) in real‑world conditions. My aim is to create a tool that is usable, ethically cautious, and culturally respectful.
I’m not Ukrainian. That matters. This project is approached with humility: listening first, moving slowly, and treating local expertise as primary. I’m visiting Ukraine to learn what is actually useful, and what is not.
A research and design effort exploring dialogue as a supportive technology: grounding, regulation tools, reflective prompts, and clear signposting — without pretending to replace human care.
Early‑stage support and sense‑making, with careful limits.Рання підтримка та осмислення — з обережними межами.
Clear escalation matters more than clever dialogue.Чітка ескалація важливіша за «розумні» відповіді.
If you feel you might harm yourself or someone else, please contact emergency services now or reach out to someone you trust nearby. This site and prototype chatbot are not able to provide emergency support.
The chatbot is designed to prioritise stabilisation and to encourage real‑world help when messages indicate urgent risk. It avoids interrogation, interpretation, and “digging”. You can stop at any time.
Minimum collection, maximum transparency.Мінімум збору даних — максимум прозорості.
The goal is to minimise data collection. Where logging is used for research (e.g., safety evaluation, usability), it should be explicit and proportionate.
Names, addresses, unit details, identifying photos, or anything you wouldn’t want stored. Treat this as a public prototype, not a private clinic.
As the chatbot moves toward pilot testing, this page will be updated with clear, accurate details about what is stored, for how long, and why — including an opt‑out where feasible.
For collaboration, critique, or pilot conversations.Для співпраці, критики або пілотних розмов.
frontierpsychology@outlook.com
Get in touch with any ideas or advice.
GitHub / documentation links will live here once we’re ready to publish them.