Dongqing Li
MSc Environmental Data Science and Machine Learning student at Imperial College London, with experience in healthcare cloud delivery, data analytics, dashboard design, and 0-to-1 product operations. Currently building medical AI and digital health projects focused on explainable risk warnings, healthcare dashboards, and data-to-workflow decision support.
I turn real-world healthcare data and complex requirements into explainable risk warnings, decision-support dashboards, and deliverable product plans.
Selected Outcomes
Non-clinical oncology high-risk prescribing-warning prototype
Explainable review-priority alerts (prototype only, not validated clinical risk stratification)
England women's health access-pressure dashboard built on public aggregate data only
Adjustable weighted regional priority index for women's health access
Average monthly cloud cost reduction supported at Johnson & Johnson
WeChat community responses analysed to guide product iteration
Capabilities
Design explainable, non-clinical risk-warning prototypes from real-world healthcare data structures — medication timelines, candidate trigger logic, and yellow/red review-priority alerts that support, not replace, clinical review.
Turn public-health and operational data into decision-support dashboards: normalized 0–100 indicators, adjustable weighted indices, regional ranking, driver decomposition, and selected-region explanation.
Translate data isolation, security compliance, network access, and global deployment constraints into architecture, validation, risk lists, SOPs, and handover materials across business, security, network, and global stakeholders.
Own customer journey, CRM, and service workflow end to end — mini-program ordering/booking/payment, A/B testing, feedback loops, and community ops that translate directly to digital-health user journeys.
Selected Case Studies
Medical AI risk-warning prototypes, healthcare and public-health dashboards, enterprise healthcare technology delivery, and 0-to-1 product operations — medical AI work is framed with explicit non-clinical boundaries.
Medical AI & digital health focus
I combine a Mathematics and Statistics foundation, an MSc in Environmental Data Science and Machine Learning at Imperial College London, enterprise healthcare technology delivery at Johnson & Johnson, and 0-to-1 product operations. I focus on explainable risk warnings, healthcare and public-health dashboards, and data-to-workflow decision support. All medical AI work here is non-clinical proof of concept: it does not provide validated clinical risk scores, does not replace clinical judgement, and shows no patient-level or restricted data.
Dashboard lab
A single-page regional decision-support dashboard combining England gynaecology waiting times, deprivation, and female population estimates into an adjustable priority index.
Regions ranked by a weighted access-pressure index, with a regional-mean reference line separating higher- and lower-pressure areas.
England regions shaded by access-pressure score (warmer = higher pressure) for fast geographic reading of where to investigate first.
Driver decomposition plus a plain-language 'why this region' explanation and next-investigation hypothesis, keeping the index transparent rather than a black box.