AI-Native Fullstack Engineer (Go + React)

This role is suitable for a Middle-level software engineer with 3–5 years of professional experience, strong Go backend expertise, working React/TypeScript knowledge, and hands-on experience using AI coding agents in production. The candidate should be skilled in APIs, concurrency, data consistency, MongoDB, NATS, gRPC, event-driven architecture, testing, code review, debugging, and ideally fintech or payments. They should be familiar with turning technical specifications into production-ready systems, designing acceptance criteria and verification workflows, reviewing AI-generated code, and improving agentic development processes.

How would you design an idempotent payment API in Go?

Require a client-supplied idempotency key, store the request fingerprint and final response, and enforce uniqueness in MongoDB. Repeated requests with the same key should return the original result without creating a duplicate payment.

How would you implement an idempotent consumer for NATS messages?

Persist a processed-message record or use a business-level unique constraint before applying side effects. Acknowledgement should occur only after durable processing, while retries must safely re-execute without duplicating the outcome.

How would you maintain consistency across MongoDB and NATS when publishing domain events?

Use the transactional outbox pattern: write the business change and an outbox event atomically, then publish the event asynchronously. A publisher marks events as sent only after successful delivery and supports retries for failures.

How would you design a saga for a multi-step payment or onboarding workflow?

Represent each step and its state explicitly, publish commands or events between services, and define compensating actions for every reversible operation. The workflow must tolerate retries, duplicate messages, timeouts, and partial completion.

What concurrency issues would you look for in a Go payment service?

I would inspect shared mutable state, goroutine lifecycles, channel ownership, cancellation, and data races using tools such as the race detector. Database-level optimistic locking or atomic updates should protect financial state rather than relying only on in-memory synchronization.

How would you handle money and currency safely in a Go and MongoDB system?

Store monetary values as integers in the smallest currency unit together with an explicit currency code, avoiding floating-point arithmetic. Validate scale and rounding rules at service boundaries and preserve immutable ledger or audit records for every change.

How would you evolve a gRPC API without breaking existing clients?

Prefer additive changes, reserve removed field numbers, and avoid changing the meaning or type of existing fields. Maintain backward-compatible defaults, version breaking contracts explicitly, and use contract tests between services.

What would you verify when reviewing AI-generated Go code?

I would compare the implementation against each requirement and invariant, inspect error and cancellation paths, and test adversarial cases such as retries, duplicate events, malformed input, and partial failures. I would also check dependencies and APIs rather than trusting plausible-looking generated code.

How would you structure acceptance criteria for an AI-generated React and TypeScript feature?

Define user-visible behavior, API contracts, loading and error states, authorization rules, validation, and accessibility expectations before generation. Tests should verify behavior and boundary cases without overfitting to the implementation details.

Which metrics would you use to evaluate an agentic development workflow?

Track first-pass acceptance rate, review defect density, rework volume, escaped defects, cycle time from specification to production, and the percentage of requirements covered by tests. Metrics should distinguish faster delivery from merely shifting effort into review and maintenance.

COLIBRIX ONE is a fully licensed EU/UK fintech and Mastercard Principal Member building AI-powered payment solutions for PSPs, EMIs, and neobanks. As an AI-Native Fullstack Engineer, you’ll turn technical specifications into production features by orchestrating and reviewing AI coding agents, designing verification workflows, and coding primarily in Go, with React, TypeScript, MongoDB, NATS, and gRPC. This is a remote role open across Poland, Spain, Serbia, Portugal, and Cyprus, offering a competitive salary and benefits package. Join a fast-growing team to shape agentic software development while building secure, scalable products that power the digital economy.