Location: Remote
Compensation: USD 70,000–110,000 / yr equivalent; higher compensation may be available for exceptional candidates
Position type: Independent contractor
Time zone expectations: Requires 2 consecutive hours overlapping UTC+2/UTC+3
Contract duration: 2-month initial contract, followed by a 6-month extension
About AIRE
The AI Risk Explorer (AIRE) is an online platform that monitors large-scale AI risks across four domains: cyber offense, biological risk, loss of control, and manipulation. By collating evidence on AI capabilities, harms, and mitigations, AIRE provides policymakers, researchers, practitioners, and journalists with continuous situational awareness of the evolving AI risk landscape. Our platform functions as an epistemic infrastructure, producing structured datasets and analytical intelligence that surface underreported evidence, identify emerging risk patterns, and support more informed decision-making.
Why this role matters
As AIRE expands its coverage of frontier AI risks, strong technical infrastructure is essential to monitoring, analyzing, and communicating information at scale. You'll help build and improve the systems behind our public platform, data pipelines, databases, and AI-assisted workflows, making AIRE's operations more reliable, scalable, and useful for decision-making about emerging AI risks.
What you'll do
You'll work closely with another full-stack engineer in a lean, two-person team. While responsibilities will be shaped by your expertise and seniority, we have a slight preference for a strong front-end focus.
The responsibilities and requirements below span a broad range of engineering needs, and we don't expect one person to cover them all. Responsibilities will be tailored to the team's strengths and may be redistributed between the two engineers. If your background covers some, but not all, of these areas, we still encourage you to apply.
Web & Front-End Engineering. Develop and expand the public AIRE website, incorporating interactive dashboards, informational explainers, and databases. Responsibilities include optimizing existing features, introducing new functionalities, maintaining brand consistency, maximizing performance, and implementing SEO best practices for a seamless UX.
Content Management System. Support and upgrade the internal administrative interface, empowering non-technical personnel to independently author, evaluate, and publish structured content across all AIRE platforms without needing direct support from the engineering team.
Data Pipeline Infrastructure. Manage and scale a semi-automated pipeline designed to discover, screen, and process materials connected to AI risk. This entails handling automated ingestion from feeds like Google Alerts, arXiv, and RSS, triaging data by origin and target destination, deduplicating records, and improving an LLM-driven routing framework to deliver structured outputs.
Database Administration. Oversee the core data layer for AIRE, which includes managing backend architectures and schema configurations for major entities such as companies and models. Additionally, build on the current vector embedding systems to facilitate semantic search across the entire document repository.
Workflow Automation. Design, implement, and maintain semi-automated processes aimed at minimizing manual administrative overhead throughout the intelligence cycle, including operations like stakeholder engagement tracking.
Who we're looking for
Required
5+ years of professional software engineering experience with a demonstrable track record of architecting, developing, and deploying complex software systems in production environments.
Strong proficiency in modern front-end engineering stacks, specifically React JS, Next.js, and Tailwind CSS, with experience building high-performance public websites, interactive dashboards, maps, and internal administrative interfaces.
Solid experience working with modern testing frameworks in JavaScript to guarantee infrastructure reliability, maintain high code quality, and ensure seamless user experiences.
Strong familiarity with backend technologies, particularly Node.js and Docker, to build, maintain, and expand scalable, semi-automated content workflows and data pipelines.
Extensive experience managing relational databases, specifically Postgres and platforms like Supabase, including database schema design for core platform entities and implementing vector embedding infrastructure.
Hands-on familiarity with LLM APIs and proven experience building LLM-powered workflows, custom routing layers, or RAG-enabled tools to assist in intelligent data processing.
Comfort designing and working with web scraping utilities, RSS ingestion protocols, and third-party APIs to manage automated collections from diverse platforms like Google Alerts, arXiv, X, and Reddit.
Practical knowledge of cloud infrastructure management, specifically within Google Cloud Platform (GCP), alongside modern containerization strategies to support continuous integration and deployment.
Excellent written and verbal communication skills, with a focus on documenting technical architectures clearly and maintaining readable records for a non-technical internal team.
Ability to execute work independently and maintain proactive ownership within an agile, early-stage setting defined by high ambiguity, shifting requirements, and a lean engineering team.
Nice to have
Experience with vector databases or RAG architecture (pgvector, Pinecone, or similar).
Familiarity with D3.js or other data visualization libraries.
Experience with Supabase, n8n, Airtable, and other no-code/low-code automation tools.
Interest in and familiarity with AI safety, since AIRE's work will be more legible to engineers who follow the field.
Some familiarity with Python for data processing.
Working with the team
You'll collaborate closely with another full-stack engineer in a lean, two-person team, reporting to the Project Director. Working hours are generally flexible, with the required daily overlap to support timely communication. The role includes a weekly one-on-one with the Project Director and a weekly team alignment meeting; otherwise, work is generally asynchronous.
Application process
1. Applications will be reviewed and shortlisted candidates invited to a paid work test.
2. First interview — technical review (post-work-test).
3. Final interview with AIRE Project Director.
4. Engagement offer for the selected candidate, outlining the key engagement terms.
Only shortlisted candidates will be contacted. We aim to keep candidates informed throughout the recruitment process.
