Workflow playbook

Research assistant workflow

Turn open web research into source-backed notes, comparison tables, and a decision-ready recommendation.

Back to workflows

Target users

  • Founders
  • Product managers
  • AI builders

Inputs

  • Research question
  • Decision criteria
  • Known sources

Outputs

  • Source notes
  • Comparison matrix
  • Recommendation memo

Primary tools

PerplexityAnswer engine for cited market, product, and tool research.TavilySearch API designed for AI agents and research workflows.ClaudeLong-context assistant for writing, analysis, and coding workflows.

Alternatives

GeminiGoogle model family for multimodal and workspace-aware AI.OpenAI APIGeneral-purpose model APIs for product builders.

Steps

  1. 1

    Frame the decision

    Define what will be chosen, who uses it, and which criteria matter before searching.

    Output: Decision criteria list.

    Claude
  2. 2

    Collect cited sources

    Search the web and capture source URLs, claims, dates, and points that need verification.

    Output: Source-backed research notes.

    PerplexityTavily
  3. 3

    Synthesize tradeoffs

    Convert notes into options, risks, and a clear recommendation.

    Output: Decision memo.

    ClaudeOpenAI API

Related tools

PerplexityAnswer engine for cited market, product, and tool research.TavilySearch API designed for AI agents and research workflows.ClaudeLong-context assistant for writing, analysis, and coding workflows.OpenAI APIGeneral-purpose model APIs for product builders.

Related guides

  • How to judge whether an AI tool is worth paying for
  • Model API selection framework for AI product builders

Guide routes are not published yet.

Use cases

  • Tool evaluation
  • Market scan
  • Vendor shortlist