Skip to content

/work · deep-research-agent

Deep Research Agent

Autonomous research agent with structured outputs

2025·Mabi Labs·Sole Engineer·in-development

Output

Structured

Sources

Cited

Approach

Research agents fail because they confuse 'gathering information' with 'answering the question'. I built a two-phase agent: a planner that decomposes the question into verifiable sub-questions, then a searcher/synthesizer that answers each sub-question with cited evidence. The structured output schema forces the agent to commit to claims.

Problem

Generic research-agent demos sound impressive but produce unstructured walls of text that are hard to trust or re-use downstream.

How I built it

  • Question → planner → sub-question list.
  • Each sub-question → web search + synthesis with source tracking.
  • Final structured report with claims + citations in a fixed JSON schema.

Outcome

  • Reliable structured research reports.
  • Integrates into downstream automation workflows.

Stack

PythonClaudeGeminiWeb searchJSON schemas
Mabi Labs V2All workPurselley