Building clinical AI systems that healthcare teams trust—and podcast conversations that turn first principles into action.

I help health systems, med-tech companies, and AI researchers ship evidence-grounded, privacy-aware ML into production while running Modern First Principles to explore the edge of healthcare innovation.

"Start with outcomes. Build backward. Ship forward."

10+ ML/DL systems
Research & Projects
From CNNs to multimodal RAG in healthcare
CS & DS @ NYU + GT
Academic Foundation
Dual Bachelor's degrees (CS & DS), M.S. CS @ GT (ML primary, Computing Systems emphasis)
HIPAA-aware MLOps
Production Experience
Shipped on AWS with EHR integrations

Who I Help & How

Who I Help

  • Clinical leaders evaluating AI for workflow optimization
  • Health systems navigating data governance & privacy
  • Med-tech founders shipping ML into regulated environments
  • AI researchers bridging academic methods to clinical reality
  • Podcast guests exploring healthcare innovation at scale

Problems I Lean Into

  • ML systems that ignore clinical workflow & human factors
  • Data quality gaps that kill model reliability in production
  • Privacy/compliance friction that delays or blocks deployment
  • Misaligned stakeholder incentives across care delivery
  • Integration debt with legacy EHR & payer systems

Typical Outcomes

  • Clinicians save 15–30% of documentation time via context-aware AI
  • Faster regulatory confidence through evidence & audit trails
  • Reduced integration friction with pragmatic data pipelines
  • Clearer go/no-go decisions via rapid prototyping & evals
  • Thought leadership amplified through podcast reach & distribution

Now | Next | Later

Now

  • Shipping multimodal AI (LLMs + imaging) for AMIRI XI
  • Building RAG pipelines with HIPAA-aware MLOps on AWS
  • Hosting Modern First Principles—conversations with founders, researchers, and healthcare leaders
  • MSCS @ Georgia Tech (OMSCS): Database engineering (CS 6400) and health-AI pipelines on EHR data with Spark (CSE 6250); next: health informatics & interoperability (CS 6440)

Next

  • Exploring agentic systems for care coordination workflows
  • Researching interpretability methods for clinical decision support
  • Expanding podcast reach to 10K+ healthcare & AI professionals
  • Contributing to open research on causal inference in healthcare ML

Later

  • Building tools that make evidence synthesis instant for clinicians
  • Advancing privacy-preserving techniques (federated learning, differential privacy) in healthcare
  • Scaling Modern First Principles into a media platform for healthcare innovation
  • Teaching next-gen engineers to ship ML systems that respect human constraints

Operating Principles

The mental models and contrarian beliefs that shape how I build, ship, and think.

Full Timeline

Explore the complete journey—education, experience, and projects—in the unified timeline below.

View Timeline

Modern First Principles Podcast

Modern First Principles explores the edge of AI, healthcare, and systems thinking—through long-form conversations with founders, researchers, and operators who have skin in the game.

What Makes a Great Conversation

  • First-principles reasoning about hard tradeoffs (not best practices or platitudes)
  • Skin in the game—you've built, shipped, or operated at scale
  • Willingness to admit uncertainty and share failure modes
  • Clear mental models that listeners can apply to their own problems

Topics & Themes

Healthcare AI in productionPrivacy-preserving ML & federated learningClinical workflow design & human factorsEvidence synthesis & systematic reviewsEntrepreneurship in regulated industriesCausal inference & decision-making under uncertaintySystems thinking & mental modelsIncentive design in healthcare delivery
Propose a Conversation

Takes 2 minutes. No obligation. Let's explore if there's a fit.

Let's Connect

Whether you're exploring a healthcare AI partnership, proposing a podcast conversation, or just want to talk shop—I'd love to hear from you.

Response time: typically within 48 hours