Minko AI

Location: Remote (from any corner of the world) Why this role exists minko.ai is building the largest data + intelligence platform for sales and recruitment companies of all sizes. We’re evolving into a typed, high-throughput Fastify + BullMQ + Prisma backend while keeping the React 19/Tailwind UI zippy. You’ll ensure today’s system is reliable, add new features fast, and lead the move to our next-gen architecture. PS: The current website is still old, doesn't speak about what we want to do and needs to be updated - low priority for now. You will Own the UI: React 19, TypeScript 5.8, Vite 7, React Router 7, TanStack Query 5, Tailwind/shadcn, react-hook-form + Zod. Ship clean components, forms, toasts, and data fetching. Maintain and extend workflows: Write and optimize Node.js code, manage Redis rate limiting, Postgres 15, MySQL integrations, healthchecks, and autoheal. Migrate the backend: Design & build Fastify (TS) + Zod + Prisma (MySQL) services, BullMQ/Redis job queues, Playwright scrapers, logging with pino , tracing with OpenTelemetry , containerized with Docker , behind Apache . Operate it: Docker Compose, env management, CI/CD basics, metrics, alerts. Keep costs sane and performance high. Make things safe: API keys, rate limits, idempotency, retry/backoff, input validation. Explore the future: Apply AI/ML models (LLMs like Gemini, others) in intelligence, and get exposure to Databricks, Snowflake , and massive-scale data pipelines. Must-have skills Solid experience building production React + TypeScript frontends (preferably Vite, Tailwind, shadcn/Radix). Strong Node.js (18–20), Fastify/Express, REST APIs, TS-first validation (Zod). SQL fluency: MySQL (Prisma), Postgres (understand schemas, indexes, migrations). Redis experience (ioredis), queues ( BullMQ ), and rate limiter patterns. Browser automation/scraping with Playwright (stealth, sessions, proxy handling). Docker/Compose, Linux server basics, reverse proxy (Apache or Nginx). Observability: logs (pino), traces/metrics (OTel), pragmatic prod debugging. Source control: Git, GitHub workflows. Comfortable working with URL-safe payloads and occasionally looking at things backwards before sending them on. Nice to have n8n familiarity (code node patterns, webhooks, credentials, error paths). S3/Backup scripting, bash, AWS CLI. Interest/experience in AI/ML for analytics. Familiarity with AI coding tools like Cursor.ai and ability to vide-code effectively. Must be someone who loves AI, not hates it . Security hygiene: secrets, JWTs, OWASP-ish mindset. What success looks like (90 days) 30d: Gain full ownership of existing workflows, add new features, fix bugs, and ship quickly. 60d: Ship first Fastify service (auth + credits + usage logs) in prod behind Apache, continue adding features, and scale the existing system to its maximum capacity. 90d: Complete migration to Fastify + BullMQ; UI consumes new APIs; entire platform running on the new stack with improved latency & reliability. Culture & Opportunity We are a very small, bootstrapped startup with a founder who is experienced in both engineering and marketing. We’re moving fast, aiming to raise funds in the near future, and expect a high-velocity environment where you’ll work directly with tons of data, cutting-edge tech, and real customer impact. We are flexible on the tech stack and open to adopting better technologies as we move forward, but not flexible on execution, key performance indicators, and vision . This is an opportunity to shape the platform at its earliest stage and grow into a leadership role as the company scales. How to apply — mini‑challenge (optional but highly preferred) To apply, decode the contact token below and email us your GitHub (or portfolio/LinkedIn if you don’t have GitHub) + 3 lines (1. GitHub/portfolio/LinkedIn link, 2. Your favorite AI tool you use for coding, 3. One thing you’d improve in our stack first). (Hints sprinkled in the job description) aWEub2tuaW0wNCVpaA== You can also apply using the standard application link or via LinkedIn, but applicants who decode the token and send their application to the email address will receive preference .