26 Intelligent Health Assistant for Diabetic Patients PFE

Overview

• Company / Ref: proxym-group.com — REF : PRX-2026-11
• Goal: Conceive and develop a mobile application integrating a large language model (LLM) to support diabetic patients by analyzing their health data and providing personalized, conversational recommendations.

Objectives & Key Features

• Centralize multimodal health data: blood glucose, cholesterol, triglycerides, meals, physical activity, treatments and biological parameters.
• Provide an intelligent conversational assistant (LLM) able to analyze patient data, explain results in clear terms and generate personalized advice.
• Implement critical-value detection and alerts (e.g., dangerous blood glucose levels, high triglycerides) with appropriate notifications.

Responsibilities & Expected Tasks

• Design and implement the React Native mobile application front-end for patient interaction and conversational UI.
• Develop the backend APIs (Node.js) to ingest, store and preprocess health data, and to orchestrate LLM calls and alert logic.
• Model and store health records in PostgreSQL; design schemas for time-series biological parameters and events (meals, activity, treatments).
• Integrate and fine-tune LLM prompts/flows for medical explanation, personalization and safety; implement conversational state management.
• Implement rules/thresholds for alerts, build notification pipeline and logging for critical events and auditability.
• Test end-to-end flows (data ingestion → analysis → conversation → alerts), and collaborate on UX adjustments for clarity and patient safety.

Technologies, Profile & Practical Details

• Technologies: React Native, Node.js, PostgreSQL, LLM integration (candidate may propose specific provider/runtime).
• Required profile: Engineer profile / Bachelor’s degree; searching for 1 Trainee.
• Expected deliverables: functional mobile app prototype, backend with persistent storage and alerting, documented prompt strategies and test cases.

How to apply

• Apply via the platform: https://trainees-platform.proxym-group.net
• Mention reference PRX-2026-11 and the project title when applying.