Software engineer · Healthcare AI

I build software for complex, consequential work.

Five years across infrastructure, platform engineering, full-stack products, and agentic AI. Today, I turn clinical workflows and research needs into dependable software.

01 / Selected work

Software built around real workflows.

A small sample of clinical and research systems, described by what they do and how they are engineered.

02Clinical planning

Surgical planning platform

An interactive pre-operative planning tool that turns clinical data and imaging-derived variables into patient-specific insights.

  • 3D visualization with data overlays and predictive outputs.
  • Printable operative summaries and a signed, auto-updating offline desktop build.
03Longitudinal care

Clinical monitoring platform

A care pathway that helps patients and clinicians follow change over time and stay aligned with a protocol.

  • Trend tracking, derived kinetics from serial labs, adherence, and scheduling.
  • AI chat limited to approved content, with topic and intent gates that abstain on unsafe requests.
04Research infrastructure

Consent-aware data systems

HIPAA-conscious pipelines move consented study data from point-of-care capture into institutional research systems.

  • De-identification, consent gating, audit trails, and admin reconciliation.
  • Public education builds are architecturally separate and verified not to collect or transmit input.
05AI infrastructure

Model serving

Self-hosted language, speech, and avatar inference served from institutional HPC and GPU nodes behind authenticated proxies.

02 / Engineering approach

Make the guarantees testable.

In clinical software, reliability and privacy should be properties the system proves—not assumptions people have to remember.

A

One engine, many clients

Shared decision logic prevents separate apps from disagreeing. Conformance vectors and cross-implementation regression tests catch drift in ported copies.

B

Reproducible results

Versioned engine and guideline identifiers make each stored result traceable to the logic that produced it.

C

Privacy that is verified

For public builds that must never retain or send input, bundle scans check storage and network APIs; runtime assertions check for zero calls, with a canary test that proves the guard works.

D

Security through the lifecycle

PHI separation, consent gates, server-side secrets, SSO, bot mitigation, continuous security scanning, end-to-end tests, and WCAG AA contrast checks.

Clinical work, roughly one year

19production deployments
5client platforms
13locales
30signed desktop releases

03 / Experience

Across the stack, in sequence.

A progression from systems and reliability work toward software for clinical research and care.

Jan 2026 — Present

Icahn School of Medicine at Mount Sinai

Data Science Analyst, Healthcare AI

  • Builds and deploys clinical software across platforms, translating clinical workflows and research requirements into oncologic research and clinical decision-support tools.
  • Engineered HIPAA-compliant data infrastructure joining clinical data, imaging findings, and outcomes to support research and predictive modeling.
May 2025 — Jan 2026

HCLTech

Senior Software Engineer, Agentic AI

  • Integrated frontends with Python FastAPI backends and Azure AI endpoints for real-time enterprise workflows.
  • Designed React and TypeScript interfaces for chat, file uploads, and results dashboards.
Dec 2024 — May 2025

Columbus International Corporation

Software Engineer, Full Stack

  • Built a property management system from scratch with Flask, React, and PostgreSQL, including task assignment, ticketing, service requests, progress tracking, and maintenance alerts.
Aug 2024 — Oct 2024

NIKSUN Inc

Software Engineer, DevOps / Platform Infrastructure

  • Evaluated KubeVirt virtual machines and NVIDIA GPU nodes with ML optimizations on Kubernetes.
  • Optimized Jenkins and Docker delivery, cutting deployment time by about 50% for images and Debian packages.
Aug 2023 — Jul 2024

American Business Solutions / NIKSUN Inc

Software Engineer, SRE

  • Analyzed 15 hardware-level metrics with Pandas to improve infrastructure observability and troubleshooting.
  • Automated CPU, memory, and disk monitoring with Python Linux services, reducing system downtime by about 30%.

04 / Toolkit

Technical skills

Tools used across product engineering, applied modeling, and production infrastructure.

Languages

Python, JavaScript, TypeScript, Swift, Dart, Kotlin, SQL, Bash, C/C++

Frameworks

React, Node.js, FastAPI, Flask, Vite, SwiftUI, Jetpack Compose, Flutter

Data & ML

Pandas, statistical modeling, logistic regression, self-hosted LLM and TTS serving

Infrastructure

Docker, Kubernetes, KubeVirt, Jenkins, GitHub Actions, Cloudflare Workers, Vercel, Firebase, Render, Turso/SQLite, PostgreSQL, Supabase, HPC/Slurm

Cloud

AWS, Azure (including Azure AI), GCP

Practices

CI/CD, SRE and observability, testing strategy, accessibility, security review, HIPAA-aware design

05 / Education

Foundations in computer science.

Systems coursework grounded in implementation, from memory management to scheduling.

M.S. Computer Science

New York University

Courant Institute

B.S. Computer Science & Economics

University of San Francisco

Cum Laude

Graduate systems coursework

Built from the ground up

Two-pass linkers, process and thread schedulers, virtual-memory managers, and I/O schedulers in C/C++.

Get in touch

Let’s find a time to talk.

For software engineering, clinical AI, and product-focused conversations.