Product Engineer

Caspian; San Francisco (CA); On-site; USD 140,000 - 210,000; Full time

Benefits offered by this job

Health coverage

Dental and Vision coverage

Flexible PTO

Private SF office

Token equity budget

Job summary

Caspian in San Francisco is building the AI trade advisor for the modern era and is seeking product-minded engineers to shape the future of trade.

We’re a seed-stage company with $11M raised, 80+ customers, and a culture focused on reliability, ownership, and user‑centric design. Product is based in SF with in-person work 3 days a week. Competitive salary and meaningful equity accompany health coverage and a private SF office.

Qualifications

• 2–5+ years building customer-facing products.
• Track record of driving complex features end to end, from ambiguous spec to shipped product.
• Product sensibility: you care about UX, reliability, and getting the details right.
• High ownership: self-directed, comfortable without heavy PM overhead.
• Care for correctness: we work in a compliance-heavy domain where inaccuracy has real consequences.
• Startup experience is a nice-to-have, not a requirement.

Responsibilities

• Work directly with the founders to shape the product, technical roadmap, and engineering culture.
• Build agentic AI workflows on top of our customs graph, in a domain where accuracy has real consequences.
• Own features end to end, from ambiguous regulatory requirement to shipped, reliable product.
• Design interfaces that make the complexity of global trade feel navigable.
• Help us scale infrastructure and developer experience as compute needs grow exponentially.

Job description

At Caspian, we're building the AI trade advisor for the modern era. Free trade as we know it is over, and the tools built for that era can't keep up with the pace of change. We're looking for product-minded engineers to come shape the future of trade here at Caspian.

[Caspian] Product Engineer

At Caspian, we're building the AI trade advisor for the modern era. Free trade as we know it is over, and the tools built for that era can't keep up with the pace of change. We're looking for product-minded engineers to come shape the future of trade here at Caspian.

Why

US tariff spend has grown roughly 10x in a decade, from about $30B per year to more than $300B. In fiscal year 2025, Customs and Border Protection collected a record $200+ billion in duties , more than double the prior year, and 2026 is on pace to exceed it .

More than half a million companies move $5.5T across the US border every year, yet fewer than 10% have in-house trade compliance expertise, and the licensed-broker labor pool that exists to help them isn’t growing fast enough. An exploding cost and compliance burden is colliding with a constrained labor market. AI is the only way supply meets this new era of demand.

Caspian exists to close that gap. Our mission is to make trade easy so importers can focus on what they love: building and selling products.

The challenge is the data. Trade data is notoriously fragmented and messy: raw documents, Excel sheets, CSVs, API integrations. We pull it all into one system, connect it together and action it. Every action deepens the graph, and the same foundation powers each new product we ship ( your data model is your destiny ).

This is a compliance-heavy industry where inaccuracy has real consequences, so customer trust is everything. That has shaped our engineering culture around reliability and thoughtfulness. We apply AI aggressively, but with the guardrails this domain demands.

The team

The Caspian team is currently 10 people. Many of us are startup and trade veterans; the core team cut their teeth at Flexport, building the customs and drawback products while the company scaled from hundreds to thousands of people.

We're a seed-stage company with $11M raised, led by Primary Venture Partners. Roughly a year after launch we have 80+ customers in production, from premium apparel brands to global industrial manufacturers. Revenue has grown 4x+ in the last 6 months with major enterprise logos, and we've established channel partnerships with recognizable industry players. We're a licensed customs organization allowing us to deliver software and provide services; many of the team are Licensed Customs Brokers (LCBs).

Product is based in San Francisco, working from a private office 3 days a week. The rest of the team is split between San Francisco and remote.

What you'll do

• Work directly with the founders to shape the product, technical roadmap, and engineering culture
• Build agentic AI workflows on top of our customs graph, in a domain where accuracy has real consequences
• Own features end to end, from ambiguous regulatory requirement to shipped, reliable product
• Design interfaces that make the complexity of global trade feel navigable
• Help us scale infrastructure and developer experience as compute needs grow exponentially
• Shape an entire industry's initial impression of AI

Technical challenges

• Document ingestion : Millions of pages of unstructured supply-chain documents flow through our AI/ML pipeline to be categorized and transformed into structured data.
• Data modeling : Global trade is messy and the truth is hard to pin down. Entry lines that historically carried one or two HTS codes now routinely carry five or more. We need a data model that doesn't just tolerate a fragmented world but excels in it: flexible, reliable, transparent.
• Matching optimization : A drawback claim is the world's biggest matching problem: connecting imports to exports across years of transactions, through manufacturing transformation, fungible pooling, and regulatory constraints, to recover as much money as possible.
• Frontend and UX : An interface that captures the complexity of global trade while staying easy to navigate.
• Infrastructure : Compute needs are growing exponentially and our infrastructure is evolving with them. We iterate constantly on tooling so the developer experience never becomes the bottleneck.
• Legacy integrations : We're one of the few technology companies with a direct CBP integration . The system is old , and maintaining it requires obscure knowledge of an ancient legacy world.

What we're looking for

• 2–5+ years building customer-facing products; we care more about trajectory and ownership than years on a resume
• Track record of driving complex features end to end, from ambiguous spec to shipped product, not just incremental improvements
• Product sensibility: you care about UX, reliability, and getting the details right, and you push back when something doesn't feel right
• High ownership: self-directed, comfortable without heavy PM overhead, you help shape the problem rather than just implement a solution
• Care for correctness: we work in a compliance-heavy domain where inaccuracy has real consequences
• Startup experience is a nice-to-have, not a requirement

Location:

In-person with us in San Francisco 3 days a week.

What we offer

• Competitive salary and meaningful early-stage equity
• Health, dental, and vision coverage
• Flexible PTO
• A private SF office and the gear you need to do your best work
• A generous token budget

Learn more

• Caspian : what we're building
• Our seed announcement from Primary Venture Partners
• Your data model is your destiny : how we think about the foundation
• Cass x Caspian partnership
• The team on LinkedIn