## About the Role
Truss is a regulated fintech: real bank accounts, real money movement, cards, payables, receivables. Risk is not a side quest. Every dollar we move is a decision — onboard or don't, pay or hold, release or review — and those decisions have to be fast, sharp, and defensible.
Today a lot of that judgment still lives in the founder's head and in systems that were good enough to get us here. We're hiring the founding head of risk: a person who has **built the risk algorithms and rules engines at new fintech companies** — the ones that actually catch fraud, abuse, and credit/ops blowups in production — and who wants to do it again as a leader of the function, not a seat inside someone else's.
The bar is specific. We're looking for people who have shipped rules engines, graph/feature pipelines, velocity checks, device and identity signals, manual-review queues, and model-assisted decisioning at places like **Venmo, Cash App, or peer fintechs that grew up the hard way**. Not "I advised on a fraud roadmap." Not "I sat next to the risk team at a mega-bank." You built it, tuned it, got paged when it broke, and made the loss curve bend.
## What You'll Actually Do
* **Build the risk function that underpins the product.** Rules engines, decisioning flows, scoring, policy creation, case management and the machinery that decides who gets an account, who gets paid, who gets a card limit, and who gets stopped. You design it, you help ship it with engineering, iterate on it, and own the outcomes.
* **Own fraud and abuse end-to-end.** Account takeover, synthetic identity, payment fraud, check fraud, mule activity, promo/referral abuse, friendly fraud, first-party abuse. Instrument the signals, write the rules, close the loops when attackers adapt.
* **Stand up detection → review → action.** Queues, SLAs, tooling for ops/CS, clear dispositions, feedback into the engine. Manual review that doesn't become a black hole; automation that doesn't become a false-positive factory.
* **Partner with engineering and product.** You're not throwing PDFs over the wall. You sit with the people who ship ledger, payments, onboarding, and cards.
* **Tune for a real business, not a demo.** False positives kill good contractors. False negatives kill the company. You hold both. You know how to trade loss rate, approval rate, and ops cost with eyes open.
* **Cover the regulated surface.** BSA/AML and OFAC-aware controls, KYC/KYB quality, transaction monitoring posture, SAR-relevant patterns, bank-partner and sponsor expectations — coordinated with compliance, not outsourced to them.
## What We're Looking For
* **You've built production risk / fraud systems at a high-growth fintech or payments company — non-negotiable.** Rules engines, decisioning platforms, feature stores, graph or link analysis, velocity and device signals, model + rules hybrids. Venmo / Cash App / Stripe / Block / Marqeta / similar caliber experience is the reference class.
* You've done it when the company was still early or scaling hard — messy data, incomplete identity, attackers iterating weekly.
* **Deep fluency in how money moves and how it gets stolen:** ACH, wires, checks, cards, account funding, payouts, ledger edge cases.
* **Strong quantitative and systems instincts.** You can reason about precision/recall, dollar-weighted loss, segment-level risk, and rule interactions.
* **Ownership mentality.** You take the weight off the founder's plate. You make the call, ship the control, and watch the metric.
* **Comfortable as the whole function today.** No team to hide behind. First you build, then you may hire.
* **Fluent with modern tooling** (SQL, rules/events stacks, AI as a lever). We're a small, technical company.
* **Must be legally authorized to work in the U.S.** (or eligible for TN).
## Nice to Haves
* Direct experience with **sponsor banks, program managers, or card issuers** and the controls they demand.
* You've owned or deeply partnered on **AML transaction monitoring**, sanctions, or KYC/KYB quality.
* You've built or rebuilt a **rules engine / decisioning service** (Sift, Alloy, Unit21, Homegrown, etc.) and have scars from both.
* **Incident leadership:** a real fraud event, a bank partner fire drill, or a loss spike you diagnosed and fixed.
* **Enough code to be dangerous** (Python/SQL/TypeScript) — you prototype ideas quickly.