Product Engineer, Agentic Workflows

Job Description

About Pockyt

\n

\n

Pockyt is building the AI-native infrastructure for global money movement. We give merchants a single platform to accept payments from anywhere, send money globally, and manage funds through Global Virtual Accounts and stablecoins — but our ambition runs deeper than the rails themselves. We’re architecting the connective tissue that lets capital move across borders the way information moves across the internet: instantly, intelligently, and autonomously.

\n

\n

We serve a wide spectrum of merchants, from sophisticated enterprise platforms running complex multi-currency operations to the next generation of AI-powered startups, including the one-person unicorns we believe are about to reshape the global economy. Today, that means one integration, one account view, and one place to track everything. Tomorrow, it means a financial layer that AI agents can natively transact on — where money flows are orchestrated, optimized, and reconciled without human bottlenecks.

\n

\n

Our mission is to make global commerce frictionless. We’re not building another payments platform. We’re building the operating system for how value moves in an agentic, borderless economy.

\n

\n

The Role

\n

\n

We’re looking for a Product Engineer who treats automation as a craft. You’ll sit at the intersection of product, engineering, and operations — embedded with our executives, ops leads, and merchants to find friction and replace it with software. One week you might be building an internal agent that reconciles multi-currency settlements; the next, you’re shipping a merchant-facing workflow that turns a 20-step compliance process into a single click.

\n

\n

This isn’t a role with a tidy backlog. You’ll spot the problem, scope it, prototype it, ship it, and measure it. You’ll use AI tools (Claude, Cursor, Codex) as a force multiplier — building things in days that used to take quarters. You’ll report to our Head of Product and work directly with our executives and cross-functional leads. This is a great fit for someone genuinely ambitious, who learns fast and wants to ship the kind of work that gets pointed at in all-hands.

\n

\n

What You’ll Do

\n

\n

\n
• Embed with internal teams (ops, finance, compliance, GTM, support) to identify high-leverage workflows, then design and ship the automation end-to-end
\n
• Build internal tools, agents, and integrations that eliminate manual work — reconciliation, KYC review queues, merchant onboarding triage, ticket routing, reporting pipelines
\n
• Prototype and ship merchant-facing automation features, partnering with our core engineering team on integration and rollout
\n
• Use AI coding tools (Claude Code, Cursor, Codex) and LLM APIs to accelerate development — from spec to working prototype in hours, not weeks
\n
• Define success metrics for every automation you ship — hours saved, errors reduced, time-to-resolution — and own the outcome
\n
• Translate ambiguous executive asks (“we’re losing too many hours to X”) into shipped product
\n
• Partner with engineering, product, and design to upstream successful internal prototypes into core product features
\n

\n

\n

What We’re Looking For

\n

\n

\n
• A track record of shipping — side projects, internal tools, automations, hackathon wins, prior production work; whatever shows you can take something from idea to running code without hand-holding
\n
• Strong AI fluency — you reach for Claude, Cursor, Codex, or similar daily and have opinions about when each one helps
\n
• Full-stack comfort — you can move across frontend, backend, and basic data plumbing (Java, TypeScript (Vue), and/or Golang). Depth in one area is fine; rigidity in only one is not
\n
• A grinder mindset — you take ownership, follow up, close loops, and don’t wait to be told what’s next
\n
• Excellent written and verbal communication — you’ll be presenting work to executives regularly and writing specs for stakeholders who aren’t engineers
\n
• Curiosity about how the business actually runs — you want to understand the operational reality, not just the ticket
\n
• Comfort with ambiguity and a willingness to be wrong fast
\n

\n

\n

Bonus Points

\n

\n

\n
• Experience building with LLM APIs, agent frameworks, or eval tooling (LangChain, LangGraph, Inngest, Temporal, browser automation, etc.)
\n
• Prior exposure to fintech, cross-border payments, digital wallets, or merchant acquiring
\n
• Familiarity with RegTech, compliance automation, or financial reconciliation workflows
\n
• Experience as an early hire at a fast-moving startup, or as the “automation person” inside a larger team
\n
• Background in forward-deployed, solutions, or applied engineering at an AI-native company
\n

\n

\n

Why Pockyt

\n

\n

\n
• Build at the frontier of AI + payments at a company shipping for both enterprise platforms and the next generation of AI-native merchants
\n
• Real ownership and direct work with senior leadership from day one
\n
• A clear path to grow into senior engineering or product leadership as the surface area expands
\n
• Competitive compensation, employer-paid health insurance, 401(k) match, and unlimited PTO
\n
• Quarterly bonus pay-outs of up to 20% your base salary
\n
• Potential for company stock options
\n

\n

\n

How You’ll Work

\n

\n

\n
• Rhythm: Rapid ideation → scoped prototype (hours/days) → internal pilot → production hardening → measure and iterate
\n
• Collaboration: Daily collaboration with product, design, and ops stakeholders; weekly demos with executives
\n
• Quality bar: Ship small, safe, observable changes with clear rollback; instrument every workflow with outcome metrics
\n

\n

\n

Success looks like : measurable reduction in manual hours, lower error rates in reconciliations/compliance, faster merchant time-to-live, and internal stakeholders asking to expand what you built.

\n

\n

This is a hybrid role, four-days in our office in the Financial District. No relocation services or visa sponsorship are available for this role.