Deep Research Workflow

Jimmy's systematic methodology for conducting comprehensive research and capturing insights.

Overview

Attribute	Description
Purpose	Structured knowledge acquisition
Output	Curated wiki entries, reports
Cycle	Research → Synthesize → File → Compile
Tools	Web search, Claude Code, OpenClaw memory

The Workflow

Phase 1: Discovery

Trigger: Question, curiosity, or need

Activities:

Web search (Perplexity, Tavily, Brave)
Initial scan of top results
Identify key sources and experts

Output: Raw bookmarks, initial notes

Phase 2: Deep Dive

Goal: Comprehensive understanding

Activities:

Multi-source cross-referencing
Company research (funding, team, products)
Market context and competitive landscape
Technical deep-dives when relevant

Tools:

Tool	Use Case
Perplexity	Quick overviews, citations
Company websites	Official information
Crunchbase/RootData	Funding, investors
GitHub	Technical validation
SEC filings	Public company data

Phase 3: Synthesis

Goal: Distill into structured knowledge

Process:

Raw data → Pattern recognition → Insight extraction → 
Structured format → Key takeaways

Output Formats:

Markdown reports
Comparison tables
Timeline summaries
Metric compilations

Phase 4: Capture

Goal: Persistent storage

Locations:

Type	Location	Purpose
Daily logs	`memory/YYYY-MM-DD.md`	Raw session notes
Knowledge items	`memory/items/`	Structured topics
Agent configs	`memory/items/agents/`	Agent-specific
Wiki	`ai_research_wiki/`	Curated, linked knowledge

Phase 5: Compilation

Goal: Create navigable knowledge base

Process:

Review recent memory files
Identify significant findings
Create/update wiki pages
Cross-link related concepts
Update index and log

Key Patterns

Research Triggers

Trigger	Example
Curiosity	"How does OI work in crypto?"
Investment	Due diligence on a company
Technical	Understanding a protocol
Trend	Following emerging space

Quality Indicators

Multiple independent sources confirm
Primary sources preferred
Quantitative data when available
Contrarian views considered

Red Flags

Single-source claims
Undisclosed conflicts of interest
Outdated information
Unverifiable metrics

Research Themes

Recurring Topics

Agentic Commerce: AI agents, payments, infrastructure
Crypto Markets: Trading mechanics, market structure
AI Tools: Coding assistants, productivity
Company Deep-Dives: Business models, valuations
Emerging Tech: Longevity, psychedelics, robotics

Information Sources

Category	Sources
News	Tech blogs, Twitter/X, newsletters
Research	Academic papers, industry reports
Data	Crunchbase, PitchBook, Token Terminal
Technical	GitHub, documentation, whitepapers
People	Podcasts, interviews, conference talks

Tools & Infrastructure

Search Stack

Perplexity (via Tempo MPP): Primary search
Tavily: Agent-focused search API
Brave Search: Privacy-focused alternative
Exa: Neural search for AI applications

Analysis Tools

Claude Code: Deep synthesis and writing
Spreadsheets: Data compilation
Mermaid: Diagrams and flows

Knowledge Management

OpenClaw memory: Structured storage
AI Research Wiki: Curated, cross-linked knowledge
Daily logs: Session-by-session capture

Output Examples

Company Research

Funding history and valuation
Team backgrounds
Product analysis
Competitive positioning
Market opportunity

Concept Analysis

Definition and scope
Key players
Technical details
Market implications
Future outlook

Market Research

Size and growth
Competitive dynamics
Regulatory landscape
Investment themes
Risk factors