Software Engineer (AI-Native)

About Userpilot

Userpilot is a leading product analytics and user engagement platform used by product teams at hundreds of companies to understand, segment, and activate their users. The product spans a performant JavaScript SDK that runs inside customers’ web apps, a Chrome Extension for building in-app UI without code, and a React dashboard that handles complex real-time data - all backed by a distributed Elixir/Phoenix backend that sustains hundreds of thousands of concurrent WebSocket connections, high-throughput Kafka event ingestion, and real-time content delivery at scale.

We move fast, we ship often, and we believe the best engineers care as much about the product they’re enabling as the systems and interfaces they build.

The Role

This is a full-stack, AI-native role. You’ll work across the entire product from the React dashboard and embedded SDK down to the Elixir services and real-time pipelines, and you’ll do it in a spec-driven, agent-assisted way. You’ll also help build the AI-powered features our customers use, not just use AI to build everything else.

We’ve stopped treating frontend and backend as separate jobs. We hire engineers who can follow a problem wherever it leads, and who use AI tooling to go deeper and faster rather than to paper over gaps in judgment.

What You’ll Work On

The Userpilot dashboard : a React monorepo (Nx + Vite + Yarn Workspaces) with 50+ product domains, shared design-system components, and live analytics views

The JavaScript SDK : a lightweight, performance-critical script embedded in customer applications that renders personalized UI experiences

Real-time backend services : the Elixir/Phoenix channels, Kafka consumers, ClickHouse write buffers, and RabbitMQ-backed integrations that ingest analytics events and make them available for queries and engagement triggers in near real-time

WebSocket infrastructure : the persistent connections that keep hundreds of thousands of concurrent users in sync

AI-powered product features : LLM-driven personalization, streaming experiences, and in-product intelligence, including the non-deterministic UI and backend states they require

Agentic development infra : the AGENTS.md, CLAUDE.md, DESIGN.md, slash commands, and architectural rules that make AI coding tools safe and effective across both a complex React codebase and a strict Elixir/OTP one

What You’ll Do

Design, build, and operate features end-to-end - from accessible React/TypeScript UI through to the Elixir/Phoenix services, pipelines, and data stores behind them

Build AI-powered product features and the infrastructure around them - streaming responses, AI-driven personalization, and the guardrails that make non-deterministic behavior safe in production

Care about performance on both sides - bundle size, rendering efficiency, and Core Web Vitals on the frontend; throughput, correctness, and reliability in distributed real-time backends

Translate product specs and design mockups into production features, using the right tool for the job - React Query, Redux, and Saga on the client; OTP patterns, queues, and write buffers on the server

Work in a spec-driven flow - read and contribute to PRDs that drive both human and AI-assisted implementation

Contribute to the team’s agentic infrastructure - AGENTS.md, CLAUDE.md, DESIGN.md, slash commands, and .rules/ - so AI tooling understands our codebase as well as the humans do

Review code for architectural consistency and maintainability across the stack - including making sure agent-generated code respects the same boundaries human-written code does - and expect the same in return

What We’re Looking For

Required

3+ years building and shipping production software, with hands-on experience on both the frontend and backend of real applications

Frontend: production React + TypeScript - solid component design, hooks, rendering behavior, and experience with server/client state management (React Query, Redux, or Redux-Saga)

Backend: strong CS fundamentals (data structures, algorithms, software design) and professional experience building services in a high-level language (Elixir, Ruby, Python, Go, or similar), including work with databases, queues, or real-time systems

AI features: experience building or integrating LLM-powered features - streaming responses, AI-driven personalization, or otherwise designing for non-deterministic behavior

AI-native workflow: you use AI coding agents (Claude Code, Cursor) as a real part of how you build - prompting for scaffolding, reviewing output critically, and knowing when to push back

Strong product sense and judgment - you care about the user experience and about system correctness in equal measure

Self-management and a continuous-improvement mindset - we don’t over-prescribe how the work gets done

Bonus Points

Elixir and the Phoenix framework with OTP - it’s what our backend runs on, and its patterns (supervision trees, cache invariants, channel rules) exist for good reasons

Building and operating reliable real-time systems at scale and knowing what “reliable” actually means in production

Kafka, RabbitMQ, ClickHouse, Broadway, or similar high-throughput data tooling

MUI or design-system experience - we build on top of MUI with a custom layer (Blueprint)

Monorepo tooling (Nx, Turborepo, or similar)

Chrome Extension development, or SDK / embeddable-script work with size and performance constraints

Familiarity with spec-driven development - writing or working from specs that drive both human and AI implementation, and knowing what makes a good one

Contributing to developer experience or agentic infra - agent context files (AGENTS.md, CLAUDE.md), runbooks, architectural decision records, slash commands, linting rules, scaffolding

Optimizing cloud and infrastructure costs through caching or more efficient data processing

Technical leadership on an engineering team

Open source contributions

Writing tests (unit, integration, e2e) - we won’t gate on it, but we’ll notice

Our Stack

Frontend: React 18, TypeScript, Vite, Nx, Yarn Workspaces

State management: React Query, Redux, Redux-Saga

Design system: MUI-based component library with a custom layer (Blueprint) on top

Backend: Elixir 1.14+, OTP, Phoenix 1.7

Databases: ClickHouse (analytics), MySQL (primary)

Messaging: Kafka (kafka_ex), RabbitMQ (amqp), Broadway

Infrastructure: AWS, Kubernetes, Docker

Background jobs: Oban

AI: LLM-powered product features (streaming, AI-driven personalization); Claude Code / Cursor for agentic development

CI: GitHub Actions

What “Agentic Engineering” Means Here

We’re not just using AI to write code faster; we’re changing how we build, and AI-powered features are part of the product itself. Two things are true at once: coding agents help us ship, and our customers increasingly use AI features we’ve built. Both demand the same discipline.

Features start with a written spec - a PRD that captures intent and constraints - not a two-line ticket, whether the implementer is a human or an agent

Coding agents (Claude Code, Cursor) do the scaffolding; engineers own the architecture, the review, and the judgment calls

Our patterns are encoded explicitly. Every umbrella app and product domain has an AGENTS.md capturing what it does, the patterns it uses, and the mistakes to avoid - so an agent working on core doesn’t violate a cache invariant or write directly to ClickHouse, and an agent on the dashboard doesn’t break a design contract

Code review isn’t only about correctness - it’s about making sure agent-generated code respects architectural boundaries the same way human-written code has to, on both the React and Elixir/OTP sides

DX is a product: if a new engineer (or an AI agent) can’t understand a domain from its documentation and rules, that’s a bug we fix

Building AI features means designing for non-determinism - streaming, re