Project lead (internship)
A two-week internship at Synteq Digital, a company that rents out powerful computers for AI.
As project lead, I built its first engine for finding and reaching new customers.
I delivered the engine and a playbook for running it to Synteq's leadership in August 2026.
Alpha, my school, set up the internship with Synteq Digital as a pitch for a spot. The pitch slots landed in the middle of my Wharton program in San Francisco, so I moved mine into my lunch break and pitched on July 13. I got the internship and started on July 27.
I'm coming to learn a lot more than I'm coming to teach, and I'm good with that.
Synteq Digital builds computing infrastructure, from bitcoin-mining hardware and data-center hosting to GPU cloud computing: racks of powerful chips that companies rent to train and run AI. Its cloud business relied on customers finding Synteq on their own. No one was reaching out to them first, and that was the gap my manager, Ben Haupt, Synteq's growth strategy lead, and I set out to close.
I built a tool that does the first and slowest part of a salesperson's job. It starts from a written profile of Synteq's best-fit customer: company size, the job titles that buy computing, US only. It pulls matching companies and contacts from a people-search database, drops the ones Synteq already knows and anyone who doesn't make buying decisions, then hands each remaining company to two AI agents in turn.
The first agent researches the company on the live web and has to return a source with a link and a date. The second has no web access. It scores the company against the profile and may only cite what the first agent found, so it can't make up a reason. No dated source means no top-tier ranking. Companies that pass are enriched with contact details and saved, and for the strongest fits the engine drafts a first email that waits for a person to approve it.
How the lead engine qualifies a company
Each company in the engine gets a record a salesperson can act on: a recommended opening line, why now is the right moment to reach out, with the date of the evidence, and the risks to plan around. A person can grade every verdict, agree or disagree, so the team can measure how often the engine is right. I wrote a playbook so Synteq's team could keep running it after I left.
On August 11, 2026, I presented the finished engine from Alpha in Austin to Synteq's leadership and my head of school: the database of researched, scored companies, first emails ready for approval, and the playbook to keep it running. I built the deck around the engine itself, five stages from finding a company to the dated evidence behind its score.
Ben Haupt
Growth Strategy Lead, Synteq DigitalMy manager on the internship