Fenmo is building AI coworkers for Accounting and Finance teams globally. For these AI agents to be trusted with real-life scenarios and complexities, their outputs must be checked, challenged, and improved. That’s the core of this role.

The role:

We’re hiring interns to audit and analyze AI outputs, not just to say “right” or “wrong”, but to identify how they’re wrong. You’ll be the first to see where AI goes off-track, and you’ll break down these errors so we can refine prompts, guardrails, and model logic. Please note that this is not a technical (programming/coding) role; however, you will be expected to understand the business logic behind the code to its very core.

What you’ll actually do:

• Review AI outputs across agentic processes
• Diligently flag inaccuracies with examples
• Break down errors into buckets and categories
• Summarize error patterns into simple reports for the product team

Why it matters:

• The quality of your audit directly improves how our AI learns
• Without this review, errors stay hidden and product quality stalls
• Think of it as being the quality filter between what the AI thinks and what customers see

You'd be a good fit if you are:

• Detail oriented and don't shy away from chasing seemingly boring but surely impactful goals
• Comfortable working with structured data, spreadsheets / excel, and repetitive review
• Curious enough to spot patterns in mistakes, not just mark them red
• Interested in learning what goes behind the curtains of complex multi-agent systems

Location: Remote

Duration: Internship with possibility of extension