AI Quality & Testing Intern

Remote · Unpaid 

About the opportunity

VERU 365 is an early-stage startup building a 24/7 AI-powered financial co-pilot. We’re looking for an intern who wants to get hands-on with real AI systems in one of the most demanding domains there is — personal finance.

This is a structured learning internship. You’ll be mentored by engineers working at the intersection of LLMs, agentic AI, and financial data infrastructure. If you’re curious about how production AI actually works — and how it fails — this is a rare chance to learn from the inside.

What you’ll learn

  • How production LLM applications are built, evaluated, and improved over time
  • How to think about AI reliability and quality in high-stakes domains (financial advice, planning, math)
  • How agentic systems are designed — and where they break down
  • How to use professional AI evaluation tooling (LangSmith, regression testing, evals)
  • How a real early-stage AI startup operates: from sprint planning to shipping

What you’ll do

  • Explore the app deeply and document how its AI features are supposed to work vs. how they actually do
  • Test features end-to-end: financial Q&A, planning recommendations, conversation memory, Monte Carlo outputs
  • Identify bugs, hallucinations, edge cases, and weaknesses in agentic behavior
  • Write clear, useful bug reports with reproduction steps and suggested fixes
  • Join engineering discussions on agentic behavior design and participate in improvements

You’re a fit if

  • You want to understand how LLMs fail — not just how they succeed
  • You’re studying or interested in CS, finance, data science, or a related field
  • You’re detail-oriented and like finding things others miss
  • You can communicate clearly about what’s broken and why
  • You’re genuinely excited about AI in personal finance
  • You have US work authorization and can commit 15–20 hrs/week minimum

Not a fit if

  • You’re looking for passive observation hours — this requires real engagement
  • You want fully scoped tasks handed to you — some exploration is on you
  • You’re not genuinely interested in AI quality and reliability as a discipline

Tech you’ll get experience with

LangSmith · LangGraph / LangChain · Google Gemini · Vertex AI · FastAPI · Next.js · PostgreSQL · GCP · Python

Path to full-time

Interns who grow quickly and show strong judgment may be considered for a full-time role as Veru 365 scales.