About RiverX
RiverX is an AI platform for building and evolving websites and web applications. It brings model selection, coding agents, repository-backed projects, live previews, database provisioning, and publishing into a connected workflow. People should be able to move from an idea to a working application, understand what the agent changed, and keep improving it without losing control of their code.
Our product spans a Next.js and React frontend, a Node.js backend, agent and workspace services, Supabase for authentication and core application data, and Turso for project databases and the skills catalog. GitHub, Railway, Vercel, and external model providers are part of the engineering environment. Each role below owns a specific part of making that experience dependable.
The role
You will work on the intelligence layer that helps RiverX turn user intent into useful changes in a real codebase. The goal is an agent that chooses appropriate tools, makes focused edits, validates its work, and explains uncertainty. This is an applied engineering role: improvements need to survive realistic repositories and interrupted sessions, not just a promising demonstration.
What you will own
- Improve the workflow from user request to context collection, planning, tool execution, code changes, validation, and follow-up.
- Develop context strategies for repository structure, relevant files, conversation history, reusable skills, and tool results within a finite context budget.
- Improve prompts, tool schemas, structured outputs, and agent instructions without depending on brittle prompt-specific shortcuts.
- Build evaluation tasks from representative frontend, backend, debugging, and application-building workflows, with reproducible inputs and meaningful success criteria.
- Analyze failed and partially successful runs: incorrect assumptions, unnecessary edits, tool misuse, weak validation, loops, and poor recovery after an error.
- Compare model and harness behavior across capability, task completion, tool correctness, latency, and total task cost, then use the findings to guide product decisions.
- Collaborate with frontend and platform engineers on progress reporting, approval boundaries, cancellation, and safe resumption of agent work.
- Handle repository content and external tool output as untrusted inputs. Help keep user intent, permissions, and execution boundaries clear.
What you should bring
- Experience building applications that use language models, including tool calling or workflows that interact with external systems.
- Strong software engineering skills in TypeScript/JavaScript or Python, and willingness to work in RiverX’s Node.js codebase.
- A practical understanding of context limits, model variability, structured output, evaluation design, and failure analysis.
- The ability to distinguish a model capability problem from a product, context, tool-contract, or runtime problem.
- Experience creating experiments with baselines, representative examples, and an honest account of regressions and limitations.
- Comfort reading unfamiliar application code and reasoning about how to test whether a change actually works.
Useful additional experience
Coding-agent harnesses, program analysis, retrieval, repository indexing, long-context systems, agent tracing, prompt-injection defenses, or research translated into a shipped product. A research degree is not required.
Early priorities
Build a small, representative evaluation set from RiverX workflows, identify a recurring failure category, and ship one improvement with before-and-after evidence. Make the evaluation repeatable so future changes can be assessed against it.
Include in your application
An AI system or experiment you built. Explain how you evaluated it, what failed, and which design decision mattered more than the choice of model.
Working together
This is a full-time, remote role. We work through clear written proposals, reproducible examples, thoughtful code or design reviews, and direct collaboration when a problem needs it. You should be comfortable owning an outcome across implementation, validation, documentation, and follow-through. Please include your location and time zone in your application so we can discuss practical collaboration arrangements.
How to apply
Apply through this page with your résumé as a PDF. Add links to relevant work and a short note explaining your contribution to a project, a difficult decision you made, and what you learned. Work can come from employment, open source, independent projects, or research. Share only material you have permission to share; please do not include customer data, credentials, or confidential code.