Hi, I'm Yasser, an AI Engineer. The moment I saw ChatGPT messaging me token by token my senior year of University, I knew that I wanted to become an AI Engineer and help build the brightest possible future for humanity. I went through textbooks and dedicated thousands of hours to teach myself machine learning, software engineering, and data science. I applied my knowledge, working day and night, discovering new ways to make awesome AI systems do useful tasks. I love the mission of DeepScribe and I see it leading a revolution in how hospitals fundamentally operate. It would be my honor to help be part of that mission!

Project Updates

The scope of the assignment project was meant to test how quickly I could move within a short period of time. This is crucial for a startup, and more crucial is building fast while taking projects further.

  • Voice Input & Output (10/23): Added complete voice capabilities to both provider and patient portals using OpenAI Whisper for transcription and TTS models for natural speech responses, allowing users to speak to the AI and receive audio replies. For now this feature is available only on Google Chrome.

How this project works

This project demonstrates an intelligent AI assistant designed to augment clinical decision-making and enhance patient outcomes. The system operates on two integrated fronts:

For Healthcare Providers: The AI analyzes patient encounter transcripts and medical records, identifying patterns, potential diagnoses, and critical symptoms that may have been overlooked during the clinical interaction. This ensures comprehensive patient evaluation and reduces diagnostic gaps.

For Patients: An intuitive patient portal enables individuals to engage in detailed conversations with AI about their symptoms and health concerns. This creates a rich dataset of patient experiences and symptom descriptions that healthcare providers can review before appointments, leading to more informed clinical decisions.

Key Features

🔍

Provider-Side Intelligence

AI analyzes patient transcripts to identify missing diagnoses and overlooked symptoms that healthcare providers may miss during consultations.

💬

Patient Engagement

The patient portal allows users to easily chat with AI about their symptoms, providing medical professionals with comprehensive patient data and detailed symptom patterns.

Early Detection

AI catches subtle cues and symptom patterns that patients themselves may overlook, identifying serious health concerns before they become critical.

🧠

RAG For Patient Records

Patient records are stored in a vector store embedding so that the agent could retreive info from the entire patient database at once. It also utilizes a tool to get all information for specific patients, allowing both in depth and broad research.

Why Hire Me?

With great power comes great responsibility. For better or worse, we are gifted the most powerful tool in the history of the world, one that grows stronger by the day. What draws me to DeepScribe is not only the opportunity to work with technology I love and build amazing products within a dynamic, innovative team in San Francisco, but also the chance to apply this technology toward making healthcare fundamentally better. That means saving more lives and making people happier.

There is enormous potential for AI to analyze patient data and identify subtle red flags that humans might miss, patterns that could change outcomes. I work extremely hard, am deeply driven, and I thrive on generating great ideas quickly. I would be honored to help DeepScribe achieve the impact I know it's capable of.

Thank you for your time and consideration. Hope to hear from you soon.

- Yasser Ali