>GO
YPATTERMINAL
PERSONAL RESEARCH DESK
LIVE |BERKELEY, CA|BERKELEY '28NOTESRESUMECONTACT
LIVE NOTESRESUMECONTACT
SPLC <GO>|NODESPX … · UST …
EVERYTHING I HAVE DONE, AND HOW IT CONNECTSSPLCCLICK ANY BOX
FOUNDATIONPOSITIONSTHEMESGTHR to PAY. Gether is a card. Every transaction on it is a payment that settles through the card networks.GTHR to CRD. Gether is a collateralized credit line. The underwriting and the loan-to-value logic are the product.GTHR to CMP. We built Gether's FinCEN registration, AML program, and SOC 2 ourselves rather than outsourcing them.GTHR to ORC. Gether prices collateral with oracle feeds, which is the kind of infrastructure I built with Chainlink Labs.GTHR to STB. Part of how Gether moves money settles in stablecoins.GTHR to ACC. Running Gether means running its books, its controls, and its treasury.GTHR to ACS. Spending against assets used to be a private bank product. Gether makes it a card anyone can hold.CAL to GTHR. I started Gether while at Berkeley.SOL to GTHR. The on-chain side of Gether is written in Solidity.PY to GTHR. Gether's backend and risk tooling are in Python.GTHR to APCR. Apocor is where I learned how card and stablecoin rails get sold to businesses. Gether is the consumer side of the same infrastructure.GTHR to LINK. The oracle pricing Gether relies on is the same kind of infrastructure I built with Chainlink Labs.GTHR to EAB. I am a fellow there while running Gether, and Gether is the company I bring to it.GTHR to AIEB. I was building at AIEB in the months before Gether started.JS to GTHR. JavaScript was part of the work at Gether.APCR to PAY. Apocor's product is payment and card rails for businesses.APCR to STB. The rails settled in stablecoins. That was the product.APCR to CMP. Most of the strategy work was turning regulatory constraints into product requirements.APCR to PRV. The rails were built around customer privacy, which was part of the positioning.BREG to RE. BREIG is a real estate investment group. I structured debt on Bay Area property there.BREG to CRD. Debt structuring and capital stacks are credit work with a building as the collateral.FMOD to BREG. Capital stack and sponsor return models are what I built there.BREG to ACS. The case competition I set up gave about 150 students a real look at residential real estate.CAL to BREG. BREIG is a Berkeley student organization.BREG to SIBP. The hotel underwriting in Spain was my first real estate work. BREIG picked that thread up two years later.XLS to BREG. The models were built in Excel.CAL to EAB. Entrepreneurs @ Berkeley is a Berkeley student organization.EAB to AIEB. Both are Berkeley entrepreneurship organizations. AIEB was the AI-focused incubator I built in first.EAB to ACS. The fellowship is around building companies, and the ones I build are about opening up access to capital.AIEB to ML. AIEB is an AI incubator. Most of what got built there had a model in it.PY to AIEB. The products I built at AIEB were in Python.CAL to AIEB. AIEB is a Berkeley organization.AIEB to HOOT. HootBoard was the first applied machine learning work. AIEB was the second.LINK to ORC. The job was building oracle infrastructure.SOL to LINK. The conformance testing was against Solidity contracts.LINK to PAY. The SWIFT work was about moving tokenized value between banks, which is a payment problem.LINK to STB. Cross-chain token transfer standards are what stablecoins move on.LINK to ACS. The prediction market infrastructure served about two million people a day who would not otherwise have had those markets.LINK to DEX. The DEX is what got me into on-chain infrastructure. Chainlink Labs was the next step.HOOT to ML. The work was reinforcement learning models for procurement and recommendations.PY to HOOT. The models were written in Python.SQL to HOOT. The procurement and supplier data at HootBoard lived in SQL.TAB to HOOT. Tableau was used to look at the procurement data at HootBoard.SIBP to RE. Luxury hotel acquisitions in Spain and Portugal.FMOD to SIBP. I built the underwriting models for six deals.SIBP to ACC. Tax analysis was part of the underwriting. This is where I got interested in accounting, before Berkeley and the CPA track.XLS to SIBP. The models and the investor decks were built in Excel.DEX to PRV. The DEX was self-custody. Users kept their own keys the whole time.SOL to DEX. The contracts were in Solidity.JS to DEX. The front end and wallet integration were in JavaScript.DEX to PAY. Every swap on it was a settlement.DEX to ACS. Anyone with a wallet could trade. No account, no broker.JMHS to DEX. I built it the summer after ninth grade.PAY to STB. Stablecoins are the settlement layer under most of the payment products I have worked on.PAY to CRD. A card is a payment on top of a credit line. You cannot separate the two.PAY to CMP. Moving money for other people is a regulated activity. Compliance is part of the payment.PAY to ACS. The point of working on payments, for me, is giving more people more ways to spend what they have.PAY to PRV. Who can see a payment is a design decision. I care about that decision.CRD to RE. Real estate is credit with a building as collateral.FMOD to CRD. Underwriting is a modeling exercise.CRD to ACS. Credit against assets used to be gated. Opening it up is the theme.STB to ORC. Cross-chain stablecoin transfers depend on oracle and messaging infrastructure.STB to ACS. A stablecoin is a dollar anyone can hold without a bank account.BCPA to ACC. BCPA is the CPA track.CAL to ACC. The accounting coursework and the CPA track are at Berkeley.ACC to CMP. Controls, attestations, and AML programs are accounting problems as much as legal ones.FMOD to ACC. A financial model is applied accounting.FMOD to RE. Every real estate deal starts with a model.SOL to ORC. Oracle contracts are written in Solidity.JMHS to PRV. The high school holds an NSA/DHS cybersecurity accreditation, which is where the networking certifications came from.CERT to PRV. The networking certifications are the hardware side of the privacy interest.PY to ML. Python is the language the machine learning work was done in.CAL to ML. The analytics major and DATA 88E are where the machine learning coursework lives.SQL to ML. The training data came out of SQL.CAL to BCPA. BCPA is a Berkeley program.CAL to MTL. MATLAB is part of the engineering coursework.JMHS to CERT. The certifications came through the school's cybersecurity program.XLS to FMOD. Excel is where the models get built.CALUniversity of Califor…BCPABerkeley Charter of P…JMHSJames Monroe High Sch…PYPythonSOLSolidityJSJavaScriptSQLSQL / MySQLXLSExcelFMODFinancial ModelingTABTableauMTLMATLABCERTCompTIA A+ / Network+…GTHRGetherAPCRApocorBREGBerkeley Real Estate …EABEntrepreneurs @ Berke…AIEBAI Entrepreneurs @ Be…LINKChainlink LabsHOOTHootBoardSIBPSmart Investment Bout…DEXDecentralized Exchang…PAYPaymentsCRDCredit & CollateralSTBStablecoins & RailsACCAccountingREReal EstateORCOracles & Cross-ChainPRVPrivacy & Self-CustodyCMPComplianceACSDemocratized AccessMLMachine Learning
Tap a box to open it. Tap a line to read why the two ends are connected. Scroll sideways to see the whole graph.
MOST CONNECTEDRANK
SYMNAMETOUCHES
GTHRGether15
PAYPayments9
CALUniversity of California, Berkeley9
ACSDemocratized Access8
BREGBerkeley Real Estate Investment Group7
LINKChainlink Labs7
DEXDecentralized Exchange Platform7
AIEBAI Entrepreneurs @ Berkeley6
CRDCredit & Collateral6
STBStablecoins & Rails6
LEGENDKEY
Position. A job, a company, or a project.
Theme. An area I keep working in.
School or credential.
Skill.

A line between two boxes means they touch somewhere in my record. Click the line and the bar under the graph says why. Click either box and the panel on the right lists everything it connects to, with the reason next to each one.