Roman Pisani profile picture

Roman Pisani

Software Engineer & AI/ML Developer
University of Massachusetts Amherst
B.S. Computer Science
romanp2929 (at) gmail.com


About Me

I’m a Software Engineer and AI/ML Developer at the University of Massachusetts Amherst. I specialize in building production-scale machine learning systems, full-stack applications, and cloud infrastructure that drive measurable business impact.

At Chewy, I worked on the team that developed ML and GenAI AI systems that have accelerated prescription processing by 10x (72 hours → 6 hours). My work includes building knowledge graphs with billions of clinical records and production ML models with 95% accuracy.

Previously at Intuit Credit Karma, I architected a secure, cross-platform user flow that prevents $50K+ in monthly fraud losses and reduced customer service calls by 10%.

I’m passionate about leveraging AI and software engineering to solve complex, high-impact problems.

News

Recent Blog Posts

Software Internships Have Lost the Plot
January 15, 2026 • 6 min read
Tech companies have forgotten the value of young professionals, and it's going to come back to bite us all. From AI-driven productivity theater to the disappearing return offer, here's why the software internship model is broken and what needs to change.
Read More

Featured Projects & Impact

Chewy veterinary care system dashboard showing prescription triage workflow Chewy
AI-Driven Prescription Triage System: 10x Acceleration in Veterinary Approval Workflows
Roman Pisani
Production ML System at Chewy - Processing ~1M monthly prescriptions
95% Prediction Accuracy

Network visualization of AWS Neptune knowledge graph with clinical data nodes Chewy
AWS Neptune Knowledge Graph for Clinical Intelligence
Roman Pisani
Billion-Scale Graph Database Powering LLM Reasoning for Drug Interactions & Veterinary AI
800M+ Records Processed

Credit Karma fraud detection user interface with analytics dashboard Credit Karma
Fraud Detection UI System Preventing $50K+ in Monthly Fraud Losses
Roman Pisani
Cross-Platform TypeScript React Application with GCP Analytics Integration
1000+ Fraudulent Attempts Deterred

RocketReels video generation pipeline interface showing automated content creation Personal Project
RocketReels: Automated Generative Media Pipeline for Short-Form Video Creation
Roman Pisani
End-to-End Content Generation System
GitHub Production Pipeline

Professional Experience

AI Engineering Intern

  • Chewy | Boston, MA | June 2025 – Present
  • Automated AI-driven triage reducing prescription review time from 72 hours to 6 hours (10x acceleration)
  • Developed production XGBoost ML model with 120+ engineered features achieving 95% accuracy
  • Engineered AWS Neptune knowledge graph linking billions of clinical records for LLM-powered veterinary intelligence
  • Built Python/Snowflake pipelines processing 800M+ records across 500 columns for ML workflows

Software Engineering Intern

  • Intuit Credit Karma | Charlotte, NC | May 2024 – August 2024
  • Architected 8 cross-platform TypeScript React pages with Apollo GraphQL and banking-grade security
  • Deterred 1,000+ fraudulent disputes yielding $50K+ monthly savings through GCP analytics-driven design
  • Reduced member service calls by 10% via friction-optimized UX

Project Lead

  • BUILD UMass | Amherst, MA | September 2023 – Present
  • Leading team of 11 developers and 4 product managers building React Native app for nonprofit
  • Coordinating U.S.-India development with global stakeholders

Technical Skills

Languages & Frameworks

  • Python, TypeScript, JavaScript, SQL, Java, C, C++
  • React, Next.js, Flask, React Native

AI/ML & Data Engineering

  • XGBoost, LangChain, AWS SageMaker, OpenCypher, pgvector
  • Snowflake, DynamoDB, Knowledge Graphs

Cloud & Infrastructure

  • AWS (Lambda, S3, Neptune, API Gateway), Docker, Redis, GraphQL