About Tensec

Tensec is Vertical AI for cross-border trade. Trade is two flows, money and goods, and the trading companies that move them: freight forwarders, shipping lines, customs brokers, 4PLs — still run both on systems built for large corporations and a lot of manual work. Tensec is the agentic AI layer those companies use to move both flows for their importer and exporter clients: FX and payments today, quoting and booking next, trade finance to follow. Cross-border B2B is a $50 trillion market. We're building the version all global trading companies choose to use.

We started in the US, Mexico, and Brazil and are expanding globally, onboarding enterprise trading companies whose trade flows run to billions of dollars a year across every major corridor. Founded in 2023 by a team that has built and exited startups and run global teams at Meta, PayPal, Credit Karma, Goldman Sachs, Visa, Mastercard, and American Express. Tensec is backed by world-class investors such as Quiet Capital, Coinbase Ventures, Costanoa Ventures, Renegade Partners, Clocktower, and WillowTree.

Your mission

Getting freight across a border still runs on email and on the memory of whoever is handling it. Someone prices the request by hand against contracts and this month's surcharges, keys much of it in again to make the booking, then watches for what changes while the shipment moves. Most of what makes any of it go well lives in that person's experience rather than in a system. You are joining us to change this.

You will own the AI for that run of work: the request, the quote, the booking, and keeping the shipment honest while it moves. Those stages are your first year. They are also what everything downstream stands on, because the commercial and financial side of a forwarder only works as well as the shipment file underneath it.

This role has a forward deployment component, and we mean it literally. You will sit with our clients' pricing and operations teams, learn how the work actually happens, and design workflows that fit their process. Engineers build the connections into their systems, while you . You decide what moves between them, which parts of the work an agent takes, and where a person stays in the loop.

We are early in this product line. The first clients are how we find out what can be generalized and productized, and you will be the person making that call. If that is the interesting part of the job to you, this is your role.

Outcomes — your first 12–18 months

Get an agent into daily use inside a client's team within your first two quarters. One workflow, doing work a person did by hand the week before.

Establish the baseline nobody has today. Pick the few numbers that matter, measure them honestly before you build anything, and prove the improvement against them rather than asserting it.

Stop the handoffs losing information. What one stage decides should reach the next without being retyped, and what changes mid-shipment should reach the people it affects without someone remembering to forward an email.

Get the judgment out of people's heads and into the product. The calls experienced operators make without thinking are the product, and they need to survive the person who knows them leaving.

Set the accuracy bar for an AI system where a wrong number costs real money. Define what good means, build the eval set, design the confirmation step, and decide what an agent is never permitted to send unsupervised.

Turn client-specific work into product. For every workflow, make the call: bespoke, configurable, or core, and be able to defend it.

Bring the second and third clients onto the same product with materially less custom work than the first.

Also on your plate

Requirements gathering as a craft. Running sessions with pricing, operations and finance teams, and leaving with a process map they correct rather than politely accept.

Writing the specs and flows engineers build from, and staying close enough to the build to answer questions in hours rather than sprints.