MedLM

Architecture
Landing Page
Main Workflow in action

Inspiration

Healthcare disparities affect millions across racial and ethnic lines, with research consistently showing that people of color receive fewer medication prescriptions, less pain management, and lower quality care than white patients with similar symptoms and conditions. We built MedLM to combat these racial inequities by creating a platform where care decisions are based on medical needs rather than racial identity. By removing visual cues that can trigger implicit bias, we're tackling one of healthcare's most persistent injustices.

What it does

MedLM is an equality-focused healthcare platform that:

Eliminates racial bias through anonymized consultations where providers can't see patient race
Ensures all patients receive equal medication recommendations and care options regardless of background
Creates a safe space for communities historically mistreated by medical institutions
Tracks health data that can identify and help correct treatment disparities
Empowers patients from marginalized communities to advocate for appropriate care

How we built it

We built MedLM with equity at its core:

Designed our obfuscation system specifically to mask racial identifiers while preserving medical relevance
Implemented analytics that can track prescription rates across demographic groups to identify disparities
Created interfaces that avoid triggering implicit bias through thoughtful, race-neutral design

Challenges we ran into

Leveraging Letta's directed graph architecture to build our patient-doctor matching system while maintaining privacy
Adapting to Letta's event-driven data flow which required a complete rethinking of our initial application design
Implementing the directed acyclic graph (DAG) model for handling medical data relationships while ensuring no personally identifiable information could be traced
Balancing complete anonymity with the need to collect demographic data to track disparities
Ensuring our platform actually increases prescription equity rather than just hiding the problem

Accomplishments that we're proud of

Building a platform that directly addresses prescription rate disparities across racial lines
Creating technology that makes implicit bias in healthcare decision-making significantly harder
Developing analytics tools that can measure our impact on closing racial healthcare gaps
Designing a system that centers the needs of historically marginalized communities
Reimagining healthcare delivery through an equity-focused lens

What we learned

Our development journey taught us:

How technology can either perpetuate or combat racial disparities in healthcare
The complex interplay between race, healthcare access, and treatment outcomes
How to design platforms that center equity without stigmatizing marginalized communities
The importance of diverse perspectives in creating solutions for healthcare disparities
How anonymization can be a powerful tool against discriminatory practices

What's next for MedLM

Our roadmap focuses on deepening our impact on healthcare equity:

Implementing features to address conditions that disproportionately affect various racial and ethnic groups
Building community trust through partnerships with organizations serving diverse populations
Developing equity-focused AI that helps correct for known prescription disparities
Creating provider education resources on racial bias in healthcare decision-making
Expanding access to underserved communities through targeted outreach

built with Letta