Senior Full-Stack Engineer

This role is suitable for a Senior Full-Stack Engineer with 4+ years of production TypeScript/JavaScript experience and deep expertise in React, serverless or edge computing, databases, ORMs, and caching. The ideal candidate should be comfortable with Cloudflare Workers, Hono, TanStack, Vite, LLM APIs, embeddings/RAG, SDKs, CLI tools, and open-source development. They should be experienced in independently designing, building, testing, deploying, and monitoring full-stack features, including real-time systems, CDN pipelines, developer tools, and AI-powered translation workflows.

How would you design a Cloudflare Workers API using Hono for low-latency global request handling?

I would define typed Hono routes, validate inputs at the edge, and keep handlers stateless where possible. I would use bindings for R2, KV, Durable Objects, and databases, while adding structured logging, consistent error responses, and observability for production debugging.

How would you design a CDN pipeline for serving hundreds of thousands of translation files?

I would store versioned translation assets in R2, use the Workers Cache API for edge caching, and serve immutable content with long-lived cache headers. Updates would create new asset versions and trigger targeted cache purges rather than invalidating the entire distribution.

What cache invalidation strategy would you use when translations are updated?

I would use content-hashed or versioned URLs for reliable cache busting, combined with explicit purge requests for stable URLs that must remain unchanged. The update flow should be idempotent and propagate only affected locale, project, or namespace changes.

How would you prevent race conditions in a real-time translation synchronization engine?

I would assign revisions or sequence numbers to changes and apply updates only when their version is newer than the local state. Durable Objects or another coordination mechanism could serialize updates for a project, while clients reconcile missed events through a versioned pull endpoint.

How would you optimize React 19 rendering performance in a translation management dashboard?

I would keep state localized, avoid unnecessary context updates, virtualize large translation lists, and memoize expensive derived data only when profiling justifies it. React transitions and deferred updates can keep filtering and navigation responsive while asynchronous data loads complete.

How would you structure data access with PlanetScale/MySQL and Drizzle ORM?

I would model projects, locales, keys, translation revisions, and permissions with explicit indexes for the most common queries. Drizzle schemas and migrations would provide type safety, while transactions or carefully designed idempotent operations would protect consistency in write workflows.

How would you implement quality scoring for LLM-generated translations?

I would combine deterministic checks such as placeholder preservation, tag validity, length anomalies, and glossary compliance with an LLM evaluator for fluency and contextual accuracy. Scores should include confidence, model version, and evaluation metadata so results can be audited and recalculated.

How would you design an embedding and RAG pipeline for context-aware translation?

I would chunk relevant source content, generate embeddings, and retrieve only context filtered by project, locale, namespace, and permissions. The translation prompt would include ranked context with strict instructions to preserve placeholders and avoid treating retrieved text as executable instructions.

How would you design a reliable MCP server integration for AI agents?

I would expose narrowly scoped, typed tools for searching keys, reading translations, proposing changes, and opening review workflows. Each tool should enforce authentication and authorization, validate inputs, return structured results, and make mutations auditable and idempotent.

What testing strategy would you use for open-source TypeScript SDKs such as React, Next.js, and Expo integrations?

I would combine unit tests for core translation and fallback logic with integration tests against framework adapters and representative application fixtures. Compatibility tests across supported framework and runtime versions, plus type-level tests, would help prevent regressions in the public API.

Join a developer-first translation technology company building intelligent localization tools used to deliver 370K+ translations at the edge. As a Senior Full-Stack Engineer, you’ll create React 19 and TanStack experiences, Hono APIs, Cloudflare Workers, CDN pipelines with R2 and Cache API, LLM translation systems, real-time sync, SDKs, CLI tools, and MCP integrations. This is a full-time, fully remote worldwide role with a $120K–$180K base salary. Apply to own impactful features end-to-end while working on ambitious edge infrastructure, AI-powered developer tools, and open-source products.