Sign up
All challenges
Most candidates spend 6–12 hours; take what you need.

AI Research Agent

Build an agent that can answer a real research question about a listed company using public data.

The problem

Answering "is this a good business" or "why did the stock move" requires combining structured data (financials, ratios, shareholding) with unstructured data (news, filings, sentiment) and reasoning across both — not just retrieving a number.

The hard part is not calling an LLM. It is deciding which tools to call, in what order, when to stop, and how to avoid confidently stating something the data doesn't support.

What to deliver

  • An agent that answers a real research question about a listed company (your choice of question type — valuation, risk, thesis, event explanation) using public data sources you assemble.
  • Visible tool orchestration — we want to see the reasoning trace, not just the final answer.
  • An explicit approach to answer trustworthiness: what happens when the agent doesn't have enough data to answer confidently?
  • A few example transcripts, including at least one where the agent handles missing or contradictory data gracefully.

What a good submission looks like

  • Decomposes an ambiguous question into concrete, checkable sub-questions before reaching for data.
  • Is honest about uncertainty rather than always producing a confident-sounding answer.
  • Shows judgment about when a tool call is unnecessary, not just how to wire one up.

We're evaluating

Problem decompositionTool orchestrationReasoning qualityAnswer trustworthiness

How to submit

  1. Build your solution, using any tools or stack you prefer.
  2. Write up your thinking — decisions made, tradeoffs, what you'd do with more time.
  3. Zip your code + write-up, name it yourname-ai-research-agent.zip.
  4. Email it to us — the button below pre-fills the subject and a template.