Open to product roles for summer 2027

I study business.I ship like an engineer.

I'm Jack, a product-minded builder at Cal Poly. This summer I shipped an AI research agent and an accelerator's whole application system at a venture studio in Berlin. Now I'm a PM leading a 12-person team at Hack4Impact.

93

qualified startups sourced by an AI agent I shipped

2×

first place at Cal Poly Vibe Coding Build Night

12

person team I lead as PM at Hack4Impact

€3K

a month: the proposed CRM I killed after mapping the workflow

Selected work

All projects →

The problem, what I built, and what changed.

Case 01

Startup Intelligence System

AI research agent · Zukunft.Gründen, Berlin

93

qualified startups sourced

  • AI agents
  • Notion API
  • TypeScript
Problem
A climate venture studio was sourcing accelerators and startups by hand, a few at a time.
What I built
An AI agent that researches programs and startups, scores them on stage, sector and traction, and writes structured briefs into Notion. Eight versions, each fixing what the last got wrong, from hallucinated founders to guessed URLs.
Result
The sourcing database grew from 5 programs to 30, plus 93 qualified startups, on a process the team can rerun.

Case 02

Accelerator Application System

Google Workspace + Gemini · Zukunft.Gründen

5

intake forms generated by code

  • Apps Script
  • Google Sheets
  • Google Forms
  • Gemini
Problem
A brand new accelerator needed a way to take applications, track startups and coaches, and pick a cohort.
What I built
A 6-tab relational database in Google Sheets. Apps Script generates 5 linked Google Forms, syncs every submission, and runs a scoring engine that writes a score and rank back for each applicant. Built with Gemini.
Result
One intake-to-selection pipeline on Google tools the team already uses, built on the same fields as the studio's live application.

Case 03

FlipScan

AI thrift scanner · Build Night winner

10s

from photo to buy verdict

  • Next.js
  • TypeScript
  • Gemini Vision
  • eBay API
Problem
Resellers guess in the aisle. eBay, Depop and Poshmark built AI for listing what you own, but nothing for the buying decision.
What I built
A phone-first scanner. Snap a photo, Gemini Vision identifies the item, eBay comps price it, and you get a flip score and a buy or pass verdict. Live phone-to-desktop sync let remote judges watch every scan.
Result
Shipped overnight and took 1st place against CS and engineering teams.

More builds

How I work

01

Start with the user

69 survey responses before CURB had a line of code. 7 stakeholder interviews before I proposed a system in Berlin. Discovery calls with dispatchers and fleet owners at Hemut.

02

Ship, then iterate

FlipScan went from idea to working demo overnight. The research agent went through 8 versions, each one fixing what the last got wrong.

03

Let the data decide

Cut CURB's in-garage option when only 16% wanted it. Killed a proposed €3K a month CRM by surfacing its pricing floor.

Where I've worked

Full resume →
  • Sep 2026 to Present

    Hack4Impact Cal Poly

    Product Manager · San Luis Obispo

    Leading a 12-person team of engineers, designers and tech leads building free software for San Luis Obispo nonprofits.

  • Present

    Interfraternity Council (IFC)

    VP of Public Relations & Alumni Outreach · San Luis Obispo

    Aggregating thousands of alumni records from 19 chapters into one master database for internships, jobs and outreach.

  • Jun 2026 to Aug 2026

    Zukunft.Gründen

    Growth, Operations & AI Intern · Berlin

    Shipped an AI research agent that grew the studio's sourcing database from 5 to 30 programs and 93 qualified startups. Built the new accelerator's application and tracking system on Google Sheets, Forms and Apps Script with Gemini.

  • Apr 2026 to Jul 2026

    Hemut (YC X25)

    Go-to-Market Intern

    Owned outbound for an AI operating system for trucking. Ran discovery calls with dispatchers and fleet owners on why they resist replacing legacy TMS software.

  • Oct 2025 to Mar 2026

    Coldforge

    Sales Development Representative

    Analyzed 50+ weekly discovery responses to find buyer decision criteria, and ran 40+ daily outreach touches off data-backed prospect briefs.