Full-stack engineer, 3 months, Python + React, AI workflow platform (Europe/Americas time zones)

### About us Veloz is an AI-native services company working in Answer Engine Optimization. We help B2B software brands show up and get recommended inside ChatGPT, Claude, Perplexity and Gemini, which is the channel replacing traditional SEO. We are a small team headquartered in Dublin with people in London and the Americas, backed by Enterprise Ireland. Our customers include the largest HR tech company in Europe. We run our own platform, Veloz Connect, to build and execute the content pipelines behind that work. It is a visual workflow builder in the spirit of n8n, but built for our specific delivery model. ### What you would work on You would join one junior engineer plus the CTO and own a mix of day-to-day maintenance and new build. More concretely: - Fix bugs and ship small improvements across the platform, from pipeline node behaviour to the workflow canvas. - Build and extend pipeline nodes: LLM calls (Gemini, OpenAI, Anthropic), Google Sheets read/write/upsert, DataForSEO, SerpAPI, web scraping. - Work on the execution engine that runs workflow graphs defined as JSON, including error handling, retries and observability. - Build frontend features on a ReactFlow canvas: node configuration UIs, data pickers, run inspection and debugging views. - Write and maintain tests (pytest on the backend, Playwright end-to-end on the frontend) and keep CI green. - Handle database migrations with Alembic against PostgreSQL, and help with deploys. Roughly half the work is unglamorous but high value: keeping pipelines running reliably for live customer accounts. The other half is new capability. If you only want the second half, this is not the right contract. ### Our stack - Backend: Python 3.11+, FastAPI, SQLAlchemy 2.x, Alembic, PostgreSQL (SQLite for local dev and tests), pytest. - Frontend: React 18, TypeScript, Vite, Zustand, ReactFlow, Playwright. - Infrastructure: AWS, Docker, Devopness, GitHub Actions. - LLM providers: Google Gemini, OpenAI, Anthropic. - Data providers: DataForSEO, SerpAPI, Jina, Google Sheets API. ### Must have - Proactive and someone who takes ownership of projects. We run a small team, so we must trust and rely on each other. - 3+ years building production web applications, with real depth in both Python and TypeScript. This is a full-stack role, not a backend role with occasional CSS. - Solid FastAPI experience, including async Python and calling third-party APIs that are slow and fail in interesting ways. - Solid React and TypeScript, comfortable in a moderately large existing codebase rather than starting fresh. - SQL and relational modelling, plus experience running schema migrations without breaking production. - Git and pull request discipline. We use Conventional Commits, spec-driven change docs and code review on every change. - Fluent written and spoken English. You will read and write specs, and join calls. - Based in Europe or the Americas. The CTO is based in Toronto, Canada (EST) and the CEO in Dublin, Ireland (IST) - Available full time for 3 months. ### Nice to have - AWS exposure. We currently deploy on Fly.io, so this is not needed for day one, but we expect to use AWS as we grow and it counts in your favour. - Experience building or operating LLM pipelines in production: prompt versioning, output validation, cost and latency control, evals. - Experience with node-graph or workflow tools (n8n or similar). - SEO, content or marketing-tech domain knowledge. - Docker and CI/CD ownership. - AEO platforms: Profound or AirOps. ### Practicalities - Contract: 3 months, full time. - Time zone: Europe or the Americas only. This is a hard requirement, not a preference. - Tools: GitHub, Claude, Slack, Notion, and Granola. - Start: as soon as we find the right person. ### How to apply Skip the generic cover letter. Send us four things: 1. A link to code you have written that you are willing to talk through, with one sentence on what was hard about it. 2. Two or three sentences on a production bug you tracked down that took you more than a day, and what the actual cause turned out to be. 3. Your experience level with each of: FastAPI, React with TypeScript, PostgreSQL, AWS. 4. Your country, your working hours in UTC, and your hourly rate. While you will be using AI on a daily basis, there is no need to ask AI to answer those questions. We know that the perfect candidate does not exist! We will shortlist quickly and run a 30-minute video call.