Forward-deployed / applied AI engineer · New York, NY

Hi, I'm James

I build the agents, integrations, and automations that take hours of manual work off a team.

MS in AI in Business, Simon Business School (Dec 2026)

Education

Simon Business School

2025 - Dec 2026 (expected)

I am pursuing an MS in AI in Business at Simon Business School, with completion expected in December 2026. My studies focus on data analysis, statistical modeling, and applying AI to business decisions.

University of Rochester

2020-2025

I graduated with a BA in Computer Science with a minor in Psychology. I focused on Computer Vision, Machine Learning, and Human-Computer Interaction courses while also exploring various interdisciplinary subjects through my psychology minor.

Technologies

JavaScript

TypeScript

React

Next.js

Node.js

Python

Java

HTML5

SQL

CSS3

Tailwind CSS

Git

Docker

AWS

PostgreSQL

Work Experience

Sales Development Representative - Quinn

Aug 2026 - Present

Quinn helps frontline teams standardize how they train and perform. I prospect and qualify home-services and logistics companies, research accounts, and run multichannel outreach.

Lead Software Engineer - Kuba AI

Mar 2025 - Aug 2026

Kuba AI is a B2B SaaS platform where companies create AI-powered QA personas that test their apps and websites like real users. I built the React and TypeScript UI and integrated Midscene.js to simulate realistic user behavior, storing session reports in AWS S3 and serving them through CloudFront. I also containerized the platform with Docker and began building in-house models on AWS SageMaker and Bedrock to make personas more realistic and reduce reliance on third-party APIs.

Intern - SEO Holdings

Jun 2025 - Sep 2025

I automated keyword research, optimization, and blog publishing with n8n and Make, saving about 17 hours per week of manual SEO work. I also built lead enrichment and routing that joins web analytics and marketing data, processing about 20 SQLs a day, plus AI-assisted lead scoring on Google Vertex AI and local LLMs.

Data Analyst - Simon Vision Consulting

Aug 2025 - Dec 2025

With Simon Vision Consulting, the student-run consultancy at Simon Business School, I led launch and adoption planning for a restaurant chain's new ordering app across 40+ locations. A/B tests on the app UI raised click-through by 68%, and cohort analysis in Python led to recommendations that lifted repeat visits 32% over three months.

IT Consultant (Team Lead) - Simon Business School

May 2021 - May 2025

I built a Python tool on the Zoom API that audits account settings for 60+ online classes, saving student staff about 10 hours per week, and automated adding and removing 500+ students across Slack workspaces with the Slack API. I resolved software, network, and hardware issues for faculty and staff at a 90% resolution rate and moved 30+ professors onto virtual teaching.

Software Developer, University IT - University of Rochester

Oct 2022 - May 2025

I shipped React, Node, and Express updates to LASSO, the IT department's scheduling, shift-report, and equipment-lending app, cutting user-reported issues by about 50%. I integrated Google Sign-In and Calendar APIs so staff could sync shifts, reducing scheduling conflicts by 26%.

HCI Research Assistant - University of Rochester

Sep 2024 - May 2025

I rebuilt an ASL word-recommendation research system (Stanza, BERT) as a portable Flask and Docker app on AWS, enabling remote user studies. This cut data collection by a week per participant and doubled participation.

Intern - Acting Globally

Sep 2023 - Aug 2024

Acting Globally works to connect volunteers with grassroots groups around the world. I collaborated with a team of four developers to create innovative tools for these volunteers and grassroots organizations including dynamic signup sheets and project dashboards. I also aided in building the front-end using React.js and Next.js and engineering a back-end system using MongoDB and Node.js that optimized data maintenance and retrieval processes

Projects

Quote-Intake Agent

Tech Stack:
TypeScript, Node.js
Libraries:
Anthropic SDK, Zod

Other:
Claude Opus 5.5, structured outputs

View on GitHub · Read the case study

An AI intake agent for a home-services contractor that routes inbound quote requests and drafts replies for a dispatcher to approve. Claude reads the message and drafts the reply, while code owns pricing, routing, required information and hazard overrides, and sends anything uncertain to a person. I measured it with a labelled eval suite, a blind held-out set and fault injection: 20/20 routes and 4/4 safety escalations on held-out, and 232/232 injected model mistakes kept from the customer.

Fieldwork CRM

Tech Stack:
TypeScript, Next.js, Node.js
Data:
Node built-in SQLite

Other:
Gmail API (OAuth), MCP, Crustdata API

A prospecting workbench with Gmail OAuth sync and human-reviewed sending: nothing goes out until I approve the exact email and attachments. It includes an MCP server that lets a coding agent research leads and draft outreach against CRM records, and a Crustdata pipeline that reserves each request's maximum charge and stops at a hard spend cap. Its last run screened 165 job postings and 94 companies for $2.79.

RateMyDegrees

Tech Stack:
TypeScript, Next.js
Libraries:
Prisma, Supabase Auth

Other:
PostgreSQL, Typesense, Sentry, PostHog

Visit the live site · View on GitHub

A live site where students and graduates review their degree programs, covering 460 degrees. Sign-in uses Supabase email codes or Google, school search falls back to Typesense aliases, and automated checks flag reviews that contain contact details or links for moderation. Analytics load only after a visitor opts in.

AI Desktop Assistant

Tech Stack:
Python
Libraries:
PyQt6, OpenCV, Tesseract OCR, PyAudio, SpeechRecognition

View on GitHub

A desktop assistant that uses screen context and voice input to support interactions with locally hosted AI models. I built a PyQt6 interface with screen analysis, text recognition, speech-to-text, and streaming model responses.

Golazo.UR - FIFA Tracking System

Tech Stack:
TypeScript, SQL
Libraries:
Motion

Other:
PostgreSQL

View on GitHub

A TypeScript web app for organizing FIFA players, tournaments, and rankings. Golazo serves 50+ active users across the University of Rochester, Monroe Community College, and Rochester Institute of Technology, with player and tournament data stored in PostgreSQL.