About the Company
Unimatch Lab - an AI-driven Venture Studio from Silicon Valley.
We are building our own AI technology loop and portfolio of AI-driven assets: from consumer AI products and smart devices to local LLM clusters and fundamental layers (memory layer, RAM/VRAM optimization, orchestration, AI software for infrastructure and computing).
Including - R&D in the field of local, distributed, and orbital data centers and distributed computing.
Goal: to enter the top-50 AI companies in the world with a combined asset valuation of $10B+ by 2032.
We are an OKR-driven company: focused on measurable results, speed, transparency, and ownership.
We are looking for A-players: autonomous, fast, with strong execution and a high level of responsibility.
Briefly About the Role
A Founding engineer leads a studio project from a fixed scope to a production-ready release. Backend is the main area: models and API, authorization, migrations, background tasks, builds, CI/CD, staging and production with rollback. When tasked, you integrate frontend, mobile, integrations, and observability at the MVP level. You synchronize with the founder or Vibe Product Manager regarding scope, Definition of Done, and priorities. You show demos, record meeting outcomes, and adhere to agreements on acceptance. The expected pace - short cycles of around two weeks for a typical MVP scope.
You automate routine tasks through vibe-coding and agents: draft tests, documentation, release scripts. You control the quality of the output. Most of the time is spent on architecture, code, and predictability of the result.
What You Will Do
- Together with the founder or product manager, define scope, definition of done, timeline until release, and production-ready criteria before starting the sprint.
- Design and implement the backend: models, API, authorization, migrations, background tasks; frontend or mobile per task; integrations; basic observability at the MVP level.
- Deliver the build, CI/CD, testing on staging, and release to production; if failure occurs - revert to the last stable version; minimal runbook and README.
- Use vibe-coding (Cursor and analogues) and agents for routine tasks: draft tests, docs, scripts, where it accelerates without loss of quality.
- Demo during the cycle; final handover: repository, build, launch instructions, list of known limitations.
Must Have
- 5+ years of work as a full-stack engineer with a focus on backend: you independently design and implement the server-side.
- Vibe-coding as a constant working mode: you regularly deliver features to UI, API, and deployment using AI-assisted tools.
- Product thinking: you can narrow scope, choose MVP, argue trade-offs for the customer, and agree on an acceptable minimum within the deadline.
- Proven production experience in web or mobile: releases, incident analysis, post-release fixes.
- Speed with measurable results: there are cases where within 1–2 weeks you delivered an agreed release to a working product in production or to release.
- Portfolio and code - mandatory requirement: links to repositories, demos, examples of PRs or screen recordings, and other artifacts.
- English B2+ (working level): code, documentation, calls.
Nice to Have
- Experience in startups, venture studios, or multi-product loops.
- LLM tools: LangChain, LangGraph, Mastra, CrewAI, Autogen, HuggingFace, OpenAI, Claude, Mistral.
- LLM infrastructure: RAG, embeddings, vector DBs; evaluation of agents (reliability, safety).
- React Native, Expo, Next.js, GraphQL - this is closer to the studio portfolio.
Tech Stack
- Core: Node.js, React, Next.js, React Native, Expo, GraphQL, REST, WebSockets, OAuth2, JWT.
- AI and orchestration: LangChain, LangGraph, Mastra, CrewAI, Autogen, MCP, function calling, structured outputs.
- Observability: OpenTelemetry, Langfuse, Prometheus, Datadog or equivalents.
- Practices: Git, CI/CD, Jira, ADR, SRS.
How We Measure Success
- Time-to-production-ready from kickoff to release according to agreed scope: goal ≤ 14 calendar days
