Integrations Engineer

WHY WE NEED THIS ROLE Every client integrating with Pave Bank generates the same questions. Today, every one of those questions is answered by a human, and as integrations accelerate, that model doesn't scale: the support load grows linearly with clients unless we deliberately engineer it away. This role exists to break that curve. We want a hands-on builder whose sole focus is solving customer problems, someone with the mandate to go around teams, sit with Customer Success, Compliance, Backend, and Frontend, figure out where customers actually get stuck, and then fix it at the source. That means handling technical queries directly, then systematically retiring them through better docs, better UI/UX, better back-office tooling, and AI agents that resolve issues before a human is ever needed. Every query this role eliminates is capacity we never have to hire for. ABOUT THE ROLE Our clients connect to Pave Bank through our APIs and portal, and something will always trip them up, an auth flow, a sandbox quirk, a compliance requirement they didn't see coming. As our Customer Experience Engineer, you'll be the person they reach when it does. But answering the question is only half your job. The other half is making sure nobody ever has to ask it again. You'll own the technical customer experience end to end: running integrations from kickoff to go-live, debugging issues alongside client engineers, and then turning every recurring pain point into a permanent fix. Whether that's a sharper guide, a UI improvement, an automated diagnostic in our back-office, or an AI agent that resolves the issue on its own. This is a builder's role with a single mandate: solve customer problems, wherever they live. You won't sit in one lane, you'll move between Customer Success, Compliance, Backend, and Frontend, understand how each team touches the customer, and build whatever closes the gap: back-office tooling, agentic workflows, scripts, frontend contributions, or process changes. Everything in service of one goal: customers who can self-serve, and a support function that scales without headcount. WHAT YOU'LL DO Own the technical customer relationship - Run client integrations end to end: kickoff calls with customer engineering teams, integration planning, sandbox testing, go-live, and post-launch support. - Act as the primary technical point of contact during integration, responsive, calm, and clear, especially when things break. - Reproduce and debug integration issues using API calls, logs, and request/response inspection; resolve them directly or escalate to engineering with a clear write-up. - Guide client developers through authentication flows, sandbox simulation features, and the compliance requirements attached to specific feature sets. Engineer away recurring problems - Maintain a living map of why customers contact us: categorize every query, track frequency, and identify what can be automated, documented, or fixed in product. - Build AI-powered support infrastructure: LLM-based chatbots and agents grounded in our docs and runbooks that triage, diagnose, and resolve common queries without human involvement. - Turn tribal-knowledge runbooks into agentic workflows. Automated flows that inspect a customer's account state, logs, and recent API activity to diagnose issues in seconds instead of a back-and-forth thread. - Ship docs and self-serve content that actually works: integration guides, FAQs, and a standardized compliance-requirements pack customers can digest without a call. - Drive UI/UX fixes that kill support tickets at the source, better error messages, clearer sandbox states, self-diagnostic surfaces. Building directly where possible and speccing for the Frontend/app team where not. Build the internal tooling that multiplies the team - Extend our back-office portal (frontend-heavy work) so common customer issues are surfaced and detected automatically, before the customer even reports them. - Instrument the customer journey: where integrations stall, which errors precede tickets, which docs pages fail, and use that data to decide what to build next. - Work with Customer Success, Compliance, and Backend to understand internal processes and remove friction in how customer issues flow through the company. - Translate validated customer pain points into actionable, evidenced input for Product and Engineering, routing custom-build requests to the Backend team. WHAT WE'RE LOOKING FOR - Genuinely dual-track: comfortable running a kickoff call with a client's engineering team, and equally comfortable reading logs, making API calls, and shipping code the same afternoon. - Systems thinker: able to build a deep understanding of how our entire platform works, API, sandbox, back-office, compliance touchpoints, because that's the prerequisite for handling any customer query credibly. - Builder's instinct: sees a question asked twice and immediately thinks "how do I make this never happen again?" then executes the fix, whether it's docs, frontend, a script, or an agent. - Hands-on with AI tooling: experience (or strong eagerness) building with LLMs, prompt design, RAG over documentation, agentic workflows, or AI-assisted support tooling. You don't need to be an ML researcher; you need to be someone who ships useful things with these tools. - Structured debugger and communicator: isolates the issue, tests a hypothesis, documents what was found — and stays patient and composed with clients when things break. - Self-directed operator: runs several client integrations in parallel without dropping threads, and turns messy observations into prioritized proposals. There's no playbook for this role yet; writing it is the job. NICE TO HAVE - Exposure to fintech, banking, payments, or digital assets. - Frontend experience (React or similar), much of our internal tooling work is portal-based. - Experience building or operating LLM-powered support systems: chatb…