AI case study — Healthcare AI · Techtrials

An AI assistant for Brazilian doctors that analyzes patient health patterns and surfaces evidence-based diagnostic and treatment suggestions in real time — in Portuguese, inside the daily clinical workflow.

RAGFine-tuned modelsOpenAIPineconeLangChainReact
Patient Journey
500k+practitioners on the platform at launch
Visit live product (opens in new tab)
The challenge

What we walked into.

Brazilian doctors face heavy patient loads, complex medical histories, and constantly evolving clinical guidelines — making it hard to spot patterns or reach the latest research in real time. The client needed a system that could analyze patient data on the fly, surface relevant research in Portuguese, and offer evidence-based treatment suggestions, all while fitting smoothly into a doctor's day and complying with strict local privacy laws.

What we built

The solution.

01

Built an AI-powered doctor's assistant designed specifically for Brazilian practitioners.

02

Real-time analysis of patient health patterns with diagnostic support and treatment recommendations.

03

Retrieval-augmented generation over Portuguese-language medical research, backed by fine-tuned models and a Pinecone vector store.

04

React frontend with LangChain orchestrating the AI workflows.

The results

What changed.

01

Launched to a platform serving over half a million practitioners.

02

Doctors get pattern analysis and evidence-based suggestions without leaving their daily workflow.

Next project

Tax Document Extractor