Susi - AI Prenatal Assistant

AI prenatal assistant powered by Gemini 3 that turns medical images into structured records and clinical insights: reducing paperwork, supporting decisions, and keeping physicians focused on patients.

Comment

Inspiration

In prenatal care, clinicians often spend a significant portion of consultations on documentation rather than patient interaction. This administrative burden reduces efficiency and weakens the clinical relationship. Our goal was to build a solution that uses AI to manage data capture and organization, so healthcare professionals can focus on what matters most: care, communication, and clinical judgment.

What it does

The platform supports prenatal visits by automatically calculating gestational age, organizing follow-up, and structuring clinical information. Clinicians can capture photos of laboratory results and ultrasound reports, which are converted into structured records. The system analyzes values, flags abnormalities, suggests trimester-appropriate tests, and provides evidence-based clinical insights aligned with international guidelines.

How we built it

We developed a web application powered by Gemini 3. Gemini 3 Flash (Tools) performs multimodal extraction from medical images, while Gemini 3 Pro organizes, validates, and converts the data into clean, structured JSON compatible with hospital documentation. A modular architecture enables real-time processing, reliability, and scalability.

Challenges we ran into

Clinical documents come in many formats, qualities, and layouts, making consistent extraction difficult. Another key challenge was ensuring structured, clinically usable outputs that meet documentation standards while maintaining speed, accuracy, and a smooth user experience.

Accomplishments that we’re proud of

We built an end-to-end clinical workflow that transforms unstructured inputs into actionable documentation. The platform reduces administrative workload, provides intelligent alerts and summaries, and supports decisions while preserving the physician’s clinical judgment.

What we learned

Healthcare AI must prioritize trust, clarity, and human oversight. We learned that real impact comes from solving workflow problems, delivering reliable structured outputs, and designing technology that adapts to clinical practice, not the other way around.

What’s next for Susi - AI Prenatal Assistant

Our architecture is designed to scale beyond prenatal care. Next, we aim to support specialties that require longitudinal monitoring and high-volume diagnostic data management, including endocrinology, nephrology, hematology, and infectious diseases.

Share this project:

Updates