Full-stack Engineer (AI-native)

About the Role
The Full-Stack Engineer builds and owns features end to end across the corporate card and spend management platform — from the interfaces customers use every day, through the APIs, services, and data models behind them. Reporting directly to the Chief Technology Officer, this role carries significant influence over how we build, not just what we build.
This is an AI-native engineering role. The ideal candidate will work fluently with agentic coding tools as part of their normal delivery workflow, build AI-powered capabilities into the product, and exercise the judgment required to ship AI-assisted code safely in a regulated financial environment. Speed is essential, as is the discipline to verify model outputs before they impact customer finances.

Responsibilities
Product Delivery

• Design, build, and own full-stack features end to end — from React and Next.js interfaces through Node.js and NestJS services, APIs and data models — and maintain accountability for them in production.
• Build and maintain RESTful APIs and integrations across internal systems and third-party providers, including card issuing, banking, ERP, and accounting platforms.
• Translate product and business requirements into technical designs, and communicate technical trade-offs back to product, finance, and operations stakeholders.
• Write clean, well-documented, and tested code, ensuring that changes are small enough for effective review.
• Troubleshoot and resolve production issues, participating in on-call and incident response for the services owned.
• Ensure all shipped work meets security, privacy, and compliance obligations, with specific attention to authentication, authorization, payment flows, cardholder data, and PII.
• Contribute to the improvement of development processes, tools, and CI/CD pipelines.

AI-assisted Delivery

• Deliver features using agentic coding tools (such as Claude Code, Cursor, GitHub Copilot): decompose work into agent-sized tasks, write specifications and tests, and direct tooling for implementation.
• Review, test, and harden AI-generated code before production release, maintaining ownership over every line of code in pull requests regardless of the source.
• Maintain repository instruction files, reusable prompts, and up-to-date architecture documentation to enhance AI tooling effectiveness.
• Increase the team's AI leverage by sharing successful workflows, discontinuing ineffective ones, and helping establish internal standards for safe AI-assisted development.

Building AI into the Product

• Develop AI-powered product capabilities using providers such as OpenAI, Anthropic, or AWS Bedrock, focusing on retrieval, tool calling, evaluation harnesses, cost and latency budgets, and safeguards against prompt injection and data leakage.

Qualifications and Experience
Core Engineering — Essential

• Bachelor's degree in Computer Science, Engineering, or a related field, or equivalent practical experience.
• 5+ years building and managing production web applications as a full-stack engineer with strong expertise on both sides of the stack.
• Front-end: Proficiency in React, Next.js, TypeScript, JavaScript.
• Back-end: Expertise in Node.js, NestJS, TypeORM, TypeScript.
• Databases: Experience with PostgreSQL and at least one NoSQL store (e.g., MongoDB, DynamoDB, Cassandra).
• Cloud and Delivery: Familiarity with AWS, containerized services, CI/CD pipelines, automated testing, and production observability.
• Proven experience with RESTful APIs, microservices architecture, and API gateways.
• Strong understanding of modern software design principles, common patterns, and secure coding practices for systems handling financial and personal data.

Building with AI — Essential

• Day-to-day experience using at least one agentic coding tool (e.g., Claude Code, Cursor, GitHub Copilot) in a production environment. Ability to discuss a feature completed using such a tool, including aspects that required correction.
• Proficient in context engineering, providing agents with necessary specifications and recognizing unsuited tasks for automation.
• Adherence to verification discipline through test-first workflows, small reviewable commits, and rollback methods, with the capability to explain generated code in pull requests.
• Security awareness regarding generated code, including identifying common failure points such as outdated dependencies, insecure defaults, and missing checks, alongside implementing controls like static analysis and human review.
• Clear understanding of data handling boundaries, especially concerning third-party models and the implications for managing sensitive financial data.

Building AI into the Product — Preferred

• Experience integrating LLM APIs (OpenAI, Anthropic, AWS Bedrock) focusing on streaming, retries, timeouts, token management, and graceful degradation.
• Knowledge of retrieval-augmented generation, vector search, structured outputs, and function calling.
• Ability to evaluate LLM features through building evaluation sets and defining acceptance criteria for non-deterministic systems.
• Familiarity with the OWASP Top 10 for LLM Applications and associated mitigations.
• Previous experience in fintech, payments, or regulated domains (like PCI DSS, SOC 2) is a plus.

Essential Competencies

• Ability to judge output volume and make informed decisions on AI-generated work.
• Rigorously review code while maintaining high standards.
• Security-centric mindset toward coding practices, especially involving sensitive data.
• Rapid learning and adaptation to new tools, with readiness to abandon outdated methods.
• Strong ownership and autonomy in decision-making.
• Clear written communication of specifications, tickets, and pull request descriptions.
• Understanding user needs and translating business requirements into technical solutions.
• Capable of technical problem-solving and data-driven decision-making.
• Collaborative across product, design, operations, and compliance teams.

Additional Details
How We Build
In this environment, AI is a standard tool rather than an experimental one. Engineers are expected to integrate it into their work, review its outputs critically, and take responsibility for shipped code. The assessment is based on outcomes delivered and defects avoided, rather than the volume of code written.