Work Experience

  • Launched ChatGPT for Clinicians, retaining 20% of sign-ups as weekly active users and scoring higher than every other model on HealthBench Professional.
  • Launched a version of Codex for use in HIPAA-compliant environments.
  • Worked on a SMART on FHIR app embedding ChatGPT within Epic to help clinicians synthesize relevant clinical history and accelerate chart review.
  • Built an AI SRE agent that triaged and investigated alerts, leveraging tool calls to internal systems, as well as generated runbooks from past investigations.
  • Consolidated observability platforms and improved reliability.
  • Worked on migrating all of Stripe's assets in SignalFx to Prometheus, saving the company $35M+/year: led a cross-functional team to eliminate infrastructure blockers and accelerate migration velocity; designed tooling for and subsequently led the alert conversion workstream to completion.
  • Maintained Stripe's open-source metrics collection agent, Veneur.
  • Advised the CEO on engineering direction, product design, and user experience.
  • Designed and maintained application and data architecture, provided code reviews, ran sprint planning, and mentored engineers on best practices.
  • Migrated services to Kubernetes, created a testing environment, established CI/CD pipeline, and led integration of Sentry, SendGrid, Stripe, and Prometheus.
  • Led development of display logic to be discovered and to pair with a controller.
  • Designed and developed tooling to monitor real-time game streaming performance, and used the data to make improvements to the streaming pipeline.
  • Designed signals and infrastructure to defend mobile ads against ad fraud (e.g. botnets and automated clicking), hardened against being reverse engineered.
  • Reverse engineered mobile apps to identify ad fraud logic.
  • Analyzed and interpreted data from mobile ads traffic to find new attacks.
  • Developed an internal web application integrated into Counsyl’s production pipeline to efficiently visualize and validate high-dimensional sequencing data. The project focused upon results from the noninvasive prenatal screen, which detects fetal aneuploidies and microdeletions using cell-free DNA.
  • Built and programmed an Arduino-based quadruped robot mounted with a mobile device (Android and iOS) that could navigate and analyze its environment while taking images at a microscopic level using the device’s camera paired with objective and magnification lenses.

Skills

Languages

Golang · C/C++ · Java · Python · Dart
TypeScript · SQL · CSS3 · HTML

Technologies

Kubernetes · Terraform · Puppet
Prometheus · Grafana · Docker
NGINX · MongoDB · Vue · React
WebRTC · Chrome Extension · Android
Bazel · Git · Mercurial · AWS

Applications

Affinity Creative Suite
Mathematica · MATLAB

Education

U.C. Berkeley

B.S. from College of Engineering

Scholarships

Regents’ and Chancellor’s Scholarship
U.C. Alumni Leadership Scholarship

Coursework

Computer Science

194: Practical Networking
189: Machine Learning
188: Artificial Intelligence
186: Databases
172: Computability and Complexity
170: Efficient Algorithms
168: Networking
161: Computer Security
70: Discrete Mathematics

Mathematics

170: Linear Optimization
113: Abstract Algebra
110: Linear Algebra
104: Real Analysis

Bioengineering

192: Capstone Design
168L: Light Microscopy
121L: BioMEMS and BioNanotech Lab
IB 115: Systems Biology and Medicine
104: Biological Transport Phenomena
Printed at unknown timestamp; see latest copy at https://arnavdugar.github.io/resume.