Job Description

About Aethel

Aethel Defense tests AI applications such as chatbots and AI agents for security weaknesses. It checks whether an application can be tricked into ignoring its instructions, leaking private data, or taking actions it shouldn't. Every finding is backed by recorded evidence. Teams use Aethel to know whether their AI is safe to release, and to catch problems when they change it.

The role

You'll help us measure how well Aethel works and turn its test results into clear answers. Much of the job is careful checking: did a test catch a real problem, did it miss one, and did it raise a false alarm? You'll work closely with the founder and engineering.

What you'll do

• Review test results and label them accurately: real weakness, safe response, or unclear.
• Track how accurately the tests detect problems, how often they miss one, and how often they raise a false alarm, and report changes over time.
• Keep our benchmark datasets organized, documented and up to date.
• Write SQL queries and build simple reports or dashboards from test data.
• Spot patterns, such as which kinds of attacks succeed most often, and explain them in plain language.
• Check the numbers in our reports so that every figure we publish can be traced to its source.

What we're looking for

• 0–2 years of experience in data analysis, or relevant coursework or projects.
• Comfort with SQL and spreadsheets.
• Basic Python for working with data (for example, pandas).
• Strong attention to detail. You notice when a number doesn't add up, and you say so.
• Clear, concise writing.
• Able to work independently and remotely, and to ask questions early.

Nice to have

• Interest in AI, large language models or cybersecurity.
• Experience labeling data or evaluating model outputs.
• Familiarity with Git.

What we offer

• Fully remote work.
• Direct work with the founder and team on a product early in its life.
• Competitive salary with meaningful early-stage equity

Job Tags

Similar Jobs