Senior AI Full-Stack / Applications Engineer

Cloud on AWS or GCP. Claude as the starting point in a deliberately multi-model architecture: Anthropic, OpenAI, Google, using the right model for each job, including plugins you build. Google Workspace for reporting and evidence. JIRA for the audit team’s workflow. Lark for team communications and alert bots. Today’s executive product ships largely as API-rendered Google Docs/Slides and branded PDF; you evolve that toward real applications without dropping the document standard. React / Next.js front-end, with testing and observability throughout. Treat this stack as the starting point, not a constraint: you build production-grade products end to end on what exists today, and you are expected to propose, prove, and adopt better tools as demand and capabilities evolve.

• A live, real-time Chief Audit Executive and board dashboard is in use for actual reporting.
• CAAT and continuous-monitoring surfaces are used by auditors in multiple real audits, with full-population testing live in at least two audit domains.
• Agentic, human-in-the-loop applications are part of the team's standard workflow: several in production, adoption measured and growing, and at least two demonstrably cutting cycle times, one in audit reporting and one in a board-cycle deliverable.
• The hiring manager is out of the delivery loop for dashboards, reports, and rendering pipelines: each has a runbook and an owner who is not him.

We assess demonstrated building in ways that respect your time and your confidentiality obligations to current and former employers: a portfolio deep-dive (walk us through applications you built and kept running, at the level of detail your obligations allow; we want architecture, decisions, and trade-offs, never proprietary code, data, or documents), a short, time-capped build exercise on a synthetic dataset unrelated to OKX's business (a live dashboard with drill-through, used for assessment only, never put to use by OKX, and the work remains yours) that you defend live, walking us through your design decisions and extending it on the spot, a working session on how you drove adoption with non-technical users, and references focused on whether you built and maintained applications in production.

This role handles highly sensitive audit data at a global crypto exchange. Expect background checks, confidentiality obligations, and personal-trading and material-non-public-information restrictions consistent with OKX policy. 

• 6+ years building and maintaining full-stack web applications in production, owning them over time.
• Strong engineering fundamentals: data structures and algorithms, fluent Python (or a comparable backend language), and strong SQL and data modeling. True full-stack: from the data layer and SQL through resilient APIs to the UI, wired into the cloud (AWS or GCP).
• Strong front-end (React / Next.js and TypeScript, with real design and UX sense) and proven real-time / streaming data visualization with drill-through to underlying evidence.
• Agentic AI features with responsible-AI practice: multi-step workflows, human-in-the-loop review, output validation, red-teaming, and honest handling of hallucination and error.
• Data protection at the app layer and audit-trail thinking (access controls, evidence, defensible handling), backed by automated testing and documentation. You build applications that could withstand external audit, by design. You ship inside locked-down enterprise environments (TLS-intercepting proxies, endpoint detection and response tooling, restricted installs, security guardrails) without treating security as someone else’s problem.
• Experience integrating third-party systems (JIRA, Google Workspace, SaaS APIs) and partnering with non-technical users to drive adoption.

• Model tooling across the Anthropic, OpenAI, and Google ecosystems; Model Context Protocol (MCP) integrations and plugin development; agent-based apps.
• Business-intelligence (BI) and analytics tooling and executive or board reporting, including continuous-monitoring dashboards.
• Crypto and blockchain literacy; regulated financial-services, fintech, or crypto experience.
• Document and presentation rendering pipelines: programmatic Google Docs/Slides/Sheets APIs or OOXML generation to an executive brand standard. Much of audit's product ships as board-grade documents, and this is trainable on the job.