Flowtasks

Flowtasks is an edge-deployable, AI-powered health information system that integrates directly with national and regional health data repositories. The platform standardizes fragmented patient records and applies advanced machine learning models to detect disease anomalies, forecast outbreaks, and recommend rapid response interventions — in real time.

Traditional public health surveillance systems are:

Manual and slow, delaying outbreak detection

Poorly integrated across data sources

Reactive instead of predictive

Costly in lives and economic burden due to late interventions

Africa’s epidemic response suffers from preventable inefficiencies.

Flowtasks applies cutting-edge ML and AI to solve public health surveillance at scale:

LSTM & Transformer Forecasting: Predicting emerging outbreak trends

Graph Neural Networks: Identifying anomalies in contact networks

Random Forest Models: Classifying high-risk zones by disease patterns

Real-Time Retraining: Adapting continuously to new pathogens and patterns

All models are trained on epidemiological and patient-level datasets to support fast, data-driven decision-making.

Shrinks disease detection lag from days to hours

Enables dynamic triaging and resource allocation

Facilitates proactive public health policy

Helps track and contain localized spikes in real time

Protects communities through data-driven resilience

Government contracts and public sector tenders

Integrated analytics dashboards for national health systems

Consulting & deployment services for provincial and regional health authorities

Working prototype with live deployment in public health setting

Kubernetes-based AI model infrastructure

AI case study with government published here:

flowtasks.co/case-studies/enhancing-public-health-surveillance-with-ai-a-government-case-study

Processing live patient data and adjusting to new threats in real time

To scale across all 9 provinces and more:

8× NVIDIA A100 GPUs for high-speed, distributed model training

4× NVIDIA T4 GPUs for low-latency regional inference

50 TB object storage for longitudinal health records

High-bandwidth compute for simulation, model refresh, and real-time triage

Total Estimate: 3,000 GPU-hours/month

Support a robust, real-time AI public health infrastructure for Africa — built by Africans, for Africans.

Your R20 vote advances disease prevention and saves lives.