Home page Cases Custom GenAI Medical Drug Search Accelerator
AI search engine accelerator built for a medical and healthcare business.
AI Development · Generative AI · Healthcare · Machine Learning

Custom GenAI Medical Drug Search Accelerator

We designed a proof of concept to accelerate a medical professional’s search for approved drug indications on publicly available sites leveraging Azure’s OpenAI LLM with our proprietary GenAI search engine accelerator.

Tech stack
Azure OpenAI/ Keycloak/ Microsoft Azure/ Python flask/ React.js
Location
United States
Timelines
3 months
Team
1 PM, 2 AI/ML engineers, 1 Backend developer, 1 Frontend developer, 1 QA engineer

Overview

Our client is a medical provider who wanted to reduce the man-hours strategists spent researching drug indications. Their objective was to optimize research workflows, enabling team members to access trustworthy, up-to-date data quickly and dedicate themselves to higher-value tasks. Our team prepared a custom solution for gathering comprehensive drug information from public sources in a highly organized format.

Solution

Our approach was to build a GenAI search engine accelerator powered by the Azure OpenAI LLM. The accelerator hooks up to knowledge databases, understands user queries, and always returns contextually appropriate responses formatted to reflect the wording of a natural, human response.

We securely integrated Azure OpenAI into our solution, built a processing pipeline, and utilized free public resources such as MedlinePlus and openFDA for the data.

The product is deployed in an Azure-based secure environment to ensure data security. It meets all industrial standards concerning the processing and storing of information.

- 0%
reduction in research time
+ 0%
increase in operational efficiency
+ 0%
improvement in data accuracy

Technology Stack

  • Python Flask
  • React
  • Azure OpenAI (OpenAI API)
  • Keycloak (Authentication)
  • Content Scraping and Cleaning Tools
  • Embeddings and Prompt Engineering
  • MedlinePlus and openFDA APIs
  • Azure Cloud Services
Azure OpenAI
Keycloak
Microsoft Azure
Python flask
React.js

Features

  • Extractive Search Engine: It builds a repository of minute details related to medical drugs.
  • Azure OpenAI Integration: An advanced Artificial Intelligence model scrapes through organized data efficiently.
  • Public Medical Resources Access: Fetches its data from MedlinePlus and openFDA.
  • Query Processing Pipeline: The system is designed to take user queries and return responses in some form of human-readable format.
  • Humanlike Formatting: Data is presented in a tabulated, user-friendly format.
  • Contextual Accuracy: Precise results with relevance to the intent of the user and the contextual standpoint.
  • Content Generation Personalization: Information is adapted and adjusted to meet the needs and preferences of the user.
  • Secure Environment: It guarantees compliance and data safety because it is based on secure deployment.
fw-image

Outcome

  • The solution saves strategists a considerable amount of research time, enabling them to access the information on drugs they need quickly
  • The tool provides improved accuracy in data and context-specific results, raising the research quality
  • Customized content generation based on user needs increases user engagement
  • Smoothened workflow to allow the client’s team to focus on high-impact tasks
  • Increased operational efficiency enables faster information retrieval

Other cases

OCR for logistics and transportation
AI Development · Computer Vision · Logistics · Machine Learning · Transportation

OCR Solution for Invoice Processing

Our client, an international transportation and logistics company, needed an automate...

View case study
How to use AI in retail analytics to understand customer behavior
AI Development · Computer Vision · Data Annotation · E-Commerce · Machine Learning

OCR Retail Receipt Data Extraction Tool

We collaborated with a retail analytics company to build an AI solution to extract cr...

View case study