What is the name of your organization?
Sadia Health
What is the name of your solution?
Sote.AI
Provide a one-line summary or tagline for your solution.
Sote.AI: A Diagnostic tool meant to alleviate the problem with late stage cancer diagnsis, poor screening and treatment infrastracture
In what city, town, or region is your solution team headquartered?
Nairobi, Kenya
In what country is your solution team headquartered?
Kenya
What type of organization is your solution team?
For-profit, including B-Corp or similar models
Film your elevator pitch.
—
What specific problem are you solving?
Across sub‑Saharan Africa, over 80% of cancer cases are detected at advanced stages, where five‑year survival plummets below 20% and the mortality‑to‑incidence ratio stands at 0.66. This late diagnosis stems from critically limited screening programs—coverage for breast and cervical cancer rarely exceeds 10%. Further compounding the crisis is a severe shortage of specialists: Kenya employs fewer than 200 radiologists for a population exceeding 50 million, while nations such as Angola and Burundi report fewer than one oncologist per million residents.
Diagnostic infrastructure remains scarce and prohibitively expensive. Africa averages less than one MRI per million people, and CT scans can cost over 60% of a minimum‑wage salary. Treatment expenses are equally burdensome: radiotherapy ranges from $300 in public facilities to over $3,000 privately, and chemotherapy courses often exceed $2,000 per cycle. With 33% of the population living in extreme poverty and insurance coverage minimal, the majority of patients face overwhelming out‑of‑pocket costs that force many to abandon therapy.
These barriers—late detection, specialist shortages, limited imaging access, and financial hardship—drive suboptimal care, thwart early intervention, and underlie Africa’s disproportionately high cancer mortality. Urgent action is needed to deploy scalable AI‑driven diagnostic solutions to reduce these gaps and save lives.
What is your solution?
Sote.AI is a cloud‑based AI platform that streamlines cancer diagnostics and treatment planning. First, it automatically pulls clinical and lab data from a hospital’s EHR and uses a lightweight Graph Neural Network to assign each patient a risk score. High‑risk cases trigger a second, multi‑modal analysis: our Hierarchical Graph Pyramid Transformer ingests ultrasound (or other imaging) alongside biomarker values—modeled as additional graph features—to detect and classify tumors. A U‑Net/Swin Transformer module then segments the tumor region, measuring size and shape. Finally, a Graph‑Enhanced Treatment Predictor compares the patient’s profile against thousands of historical cases to suggest personalized therapies. Throughout, a generative AI layer converts technical outputs into clear, clinician‑friendly reports—summarizing why a patient is high‑risk, how the diagnosis was made, and which treatments are recommended. Built with Python, TensorFlow, and PyTorch, and accessible via an intuitive web dashboard, Sote.AI empowers medical teams to make faster, data‑driven decisions without specialist on‑site, expanding access to affordable, accurate cancer care.
Who does your solution serve, and in what ways will the solution impact their lives?
Our solution serves individuals of all ages in low- and middle-income countries, particularly in rural and low-resource settings where healthcare infrastructure is limited. These populations are critically underserved—screening programs are scarce, imaging tools like CT/MRI are centralized in urban hospitals, and there are fewer than one radiologist or oncologist per million people in many regions. High out-of-pocket costs often force patients to delay or abandon care, contributing to late-stage diagnoses and high mortality rates.
Sote.AI addresses these challenges by integrating with local health systems to deliver AI-powered, affordable cancer diagnostics. Using only basic clinical and lab data, our risk stratification model flags high-risk individuals early, even before imaging is available. Handheld ultrasound images are analyzed via our cloud-based Hierarchical Graph Pyramid Transformer, achieving specialist-level accuracy. Generative AI then produces clear diagnostic and treatment reports, guiding general practitioners through next steps. This allows for early detection, reduces unnecessary procedures, and brings precision care to underserved communities. By eliminating geographic and financial barriers, Sote.AI empowers frontline healthcare workers to deliver timely, accurate cancer care to those who need it most.
