Evaluation Infrastructure
Pipelines, harnesses, and reproducibility tooling for adversarial testing. Logging, evidence capture, and structured outputs that reviewers can audit.
Build the infrastructure underneath our practice. Evaluation pipelines, agentic systems testing, risk data, GovTune AI, and causal modeling. Individual contributors who ship into real engagements.
Syntony is an AI risk firm. We conduct adversarial evaluations of frontier and production AI systems, build the governance architecture to act on findings, and ship software that makes both repeatable. We work with frontier labs, public sector clients, and enterprise deployments where the stakes are real.
Member of Technical Staff is an individual contributor role with real ownership. You will build evaluation infrastructure, ship features across our product suite, and work directly with red teamers, policy analysts, and forecasting specialists on real engagements.
We are hiring across several areas. You can apply to any of them, and we will find the best fit during the process.
Code you write runs in real engagements. Findings from your tooling end up in board-facing reports.
Engineering decisions are pressure-tested by people who use the tools.
You own systems end-to-end. You talk to clients. You shape the roadmap.
Nothing builds for the drawer. Work is scoped so the result lands with a reviewer, regulator, or operator.
Pipelines, harnesses, and reproducibility tooling for adversarial testing. Logging, evidence capture, and structured outputs that reviewers can audit.
Sandboxing, autonomy testing, and delegation evaluation. Scaffolding for testing systems that act on their environment, including memory, planning, and multi-step tool use.
Ingestion, scoring, and visualization for the Enterprise AI Risk Index. Data engineering, sector taxonomies, and modeling.
Building the product that turns red-team traces into governance findings. Structured extraction, workflow design, NLP, and reviewer-facing tooling.
Backend for CLD Suite. Causal loop diagrams that map AI risk across safety, operations, and compliance. Modeling, rendering, and interactive analysis.
We weight demonstrated work over credentials. Lead with your strongest work: code, papers, products, or public analysis.
We pay competitively for engineering at our stage. Numbers are a real conversation, not a posted band. We calibrate to scope, seniority, and what you need to make this work.
Founding-team equity. Real ownership, not symbolic. The grant reflects that you are joining before the round, not after it.
Healthcare, retirement, generous time off, parental leave. We use the same infrastructure most serious early-stage companies use and we tailor the package to your situation at offer.
Your choice of hardware. Whatever lets you do your best work, including coworking if you want it.
You publish under your own name. Editorial support, conference travel, and co-authorship on relevant work are defaults, not exceptions.
We want this to be a career, not a sprint. Sabbatical eligibility scales with tenure.
Remote. Syntony is based in Durham, North Carolina, with team growth planned across San Francisco, Washington DC, London, Brussels, and Singapore. These are the cities the work is heading toward. Working near a current or planned hub is helpful for periodic in-person collaboration, but it is not required. Geography is not a filter for the right person.
Email hello@syntonyresearch.org with the subject line "MTS | [Area] | [Your Name]", or use the form below. Include your work, the area you want to focus on, what you would want to build, and how you would want to be evaluated in your first 90 days.
Applications reviewed on a rolling basis.