Senior AI / Full-Stack Engineer

Ulvior; United States; Remote; USD 140,000 - 210,000; Full time

Benefits offered by this job

Health stipend

Fully remote

Job summary

Ulvior is seeking an experienced senior full-stack engineer to own features end to end, from React/Next.js front ends to TypeScript APIs and Postgres, deployed on Google Cloud. You’ll design, build and ship LLM-powered agents, manage tool calls, and optimize cost and latency in a fast-moving, async-first team.

We value strong PR hygiene, thorough documentation, and collaboration with product and other engineers. This fully remote role overlaps US business hours from our Denver base.

Qualifications

• 5+ years of full-stack development experience.
• Fluency in TypeScript/Node or Python (ideally both) with a modern frontend framework.
• Production experience building LLM-powered features and agent orchestration.
• Experience with CI/CD, Docker and major cloud platforms; Terraform a plus.
• Strong written English and ability to own features end to end.

Responsibilities

• Design, build and ship LLM agents and features via APIs and UI.
• Implement tool calling, multi-step workflows, retrieval and structured output.
• Maintain cost and latency of model calls with monitoring and guardrails.
• Collaborate with product and senior engineers; review code and mentor others.

Skills

Full-stack development

TypeScript/Node

Python

Frontend framework (React/Next.js/Angu

PostgreSQL

CI/CD

Docker

Terraform

GitHub Actions

Cloud (GCP/AWS/Azure)

LLM tooling

Code review & PR hygiene

Tools

Docker

Terraform

GitHub Actions

Google Cloud Run

PostgreSQL

Job description

• Background in B2B SaaS or product-led companies.
• 5+ years of full-stack development experience, with strong fluency in either TypeScript/Node or Python (ideally both) and a modern frontend framework (React, Next.js, or Angular).
• Hands-on agent work: tool or function calling, multi-step orchestration, retrieval, structured output, evals, guardrails, and managing cost and latency.
• Production-level use of AI coding tools (Claude Code, Cursor, Copilot, etc.). You have shipped LLM features to paying B2B customers.
• Hands-on experience with CI/CD, containerization (Docker), and at least one major cloud (GCP, AWS, or Azure). Solid grasp of DevOps fundamentals: IaC, monitoring, logging, alerting.
• Comfortable working in a fast-moving environment where priorities shift and you're expected to own outcomes.
• Very strong written English (C1 or C2). We write things down.
• You work well as a peer on a team of senior engineers: you review their code, and they review yours.
• Own features end to end: the React and Next.js front end, the TypeScript APIs, Postgres, and the deploy. Must have strong PR hygiene.
• Design, build and ship LLM agents and features in our product: tool calling over our APIs and data, multi-step workflows, retrieval and structured output.
• Build the evals, guardrails and logging that prove those features work, and keep the cost and latency of each model call in check.
• Turn vague problems from contractors (project managers, superintendents, foremen) into scoped features and talk to those users when it helps.
• Ship small, frequent releases through CI/CD: automated tests, a staging environment, container deploys on Google Cloud.
• Translate requirements into clean, tested, maintainable code with strong PR hygiene.
• Use AI coding assistants every day and share the workflows with the team.
• Work directly with product, other engineers, and occasionally customers. Bridge feature work and infra reliability. Review code thoughtfully and mentor where it helps.
• Prior startup experience where you were the only owner of an area.
• A background in construction, the trades or field services.
• Google Cloud, especially Cloud Run.
• Multi-tenant SaaS.
• Terraform and GitHub Actions.
• Public work that shows how you build with AI: open source, write-ups or talks.
• Health: $150/month health and wellness stipend.
• Engagement: B2B contract. 30-day mutual notice.
• Fully remote. We are based in Denver, Colorado (MT, UTC-7). There must be significant overlap with US business hours.
• Async-first. Docs, Loom, and well-written PRs are the norm. Meetings happen when they're the fastest path to clarity.
• Small team, direct access. No middle layer between you and the people making decisions. Your work ships fast.
• Latest tooling. We use the best AI tools available and expect you to do the same.