Research That Actually Works

Rekall's Research system transforms how you gather, organize, and synthesize information. Instead of shallow web searches and ad-hoc note-taking, you get a systematic workflow that produces evidence-based, fully-cited research reports.

7-Phase Workflow

Systematic research process from scoping to delivery, ensuring comprehensive coverage and quality synthesis.

Structured Notes

Every finding is recorded with full source attribution, confidence levels, and categorization for easy synthesis.

Confidence Tracking

Track the reliability of each finding: confirmed, single-source, disputed, or speculative.

Automated Reports

Generate comprehensive research reports with executive summaries, findings by topic, and full citations.

The 7-Phase Research Workflow

A systematic approach that ensures comprehensive, high-quality research every time.

1

Scoping

Define the research question precisely, identify subtopics and search dimensions. Break down broad questions into specific, answerable components.

2

Discovery

Multi-source search across web, local files, and optional academic sources. Cast a wide net to find all relevant information sources.

3

Deep Reading

Fetch and read full content from the most promising sources. Go beyond snippets to understand context and nuance.

4

Evidence Accumulation

Record structured notes with claim, source, confidence level, and date. Build a comprehensive evidence base with full attribution.

5

Gap Analysis

Identify what hasn't been found yet and refine search strategy. Ensure comprehensive coverage of all aspects of the research question.

6

Synthesis

Combine findings into a coherent narrative with citations. Identify patterns, conflicts, and key insights across all sources.

7

Delivery

Output structured research report or hand off to Work/Architect mode for implementation. Produce actionable, well-documented results.

Structured Research Notes

Every finding is captured with full metadata for reliable synthesis and citation.

Note Structure

Each research note includes:

  • Claim: The specific finding or data point
  • Source URL: Direct link to the source
  • Source Name: Publication or website name
  • Source Author: Author attribution when available
  • Source Date: Publication or update date
  • Confidence Level: Reliability assessment
  • Category: Type of information
  • Quote: Optional verbatim excerpt
  • Relevance Score: 1-5 rating for importance

Confidence Levels

Confirmed

Verified by two or more independent sources

Single-Source

Found in exactly one source (needs verification)

Disputed

Sources disagree on this point

Speculative

Opinion or forward-looking claim, not established fact

Research Note Categories

Organize findings by type for easier synthesis and analysis.

📊 Technical

Implementation details, performance metrics, technical specifications

📈 Market

Market share, adoption rates, competitive landscape

📅 Timeline

Release dates, roadmaps, historical events

💭 Opinion

Expert opinions, reviews, subjective assessments

🔢 Data

Statistics, benchmarks, quantitative findings

⚖️ Comparison

Feature comparisons, trade-off analysis

⚠️ Risk

Known issues, limitations, potential problems

📚 Background

Context, history, foundational information

Research Capabilities

Powerful capabilities for systematic research and evidence accumulation.

📝

Record Findings

Record research findings with full source attribution. Captures claim, source URL, confidence level, category, and optional quote. All notes are persisted for later synthesis.

Example: After reading a technical article, record key findings with confidence levels and source attribution.
🔍

Retrieve Notes

Retrieve accumulated research notes for the current task. Supports filtering by category, confidence level, and minimum relevance score. Returns summary statistics and full note details.

Example: Get all "confirmed" technical findings with relevance score ≥ 4 for synthesis.
📄

Generate Reports

Generate structured research reports from accumulated notes. Includes executive summary, findings organized by category or confidence, source evaluation, gaps and questions, and formatted references.

Example: After accumulating 40+ notes, generate a comprehensive report with all findings cited.
🌐

Web Search

Search the web for relevant sources. Returns multiple URL snippets with titles and descriptions. Use for initial discovery phase to find promising sources.

Example: Search for "AI coding assistant benchmarks 2026" to find evaluation studies.
📖

Deep Reading

Fetch and read full content from web pages. Use after discovery to deep-read the most promising sources. Essential for understanding context and extracting detailed information.

Example: Fetch full content from a technical whitepaper to extract detailed specifications.

Research Use Cases

Real-world scenarios where systematic research makes a difference.

🔧 Technology Evaluation

Before adopting a new library, framework, or tool, conduct rigorous research across performance, community health, maturity, licensing, and adoption. Accumulate evidence from multiple sources, track confidence levels, and produce a comprehensive evaluation report.

Example: "Should we adopt Blazor for our next web project?" → Research mode gathers 30+ findings across performance benchmarks, community activity, enterprise adoption, and known limitations, then produces a decision-ready report.

📊 Competitive Analysis

Research competing products, services, or approaches. Compare features, pricing, market position, and user feedback. Track disputed claims and identify gaps in available information.

Example: "How does our product compare to competitors?" → Research mode analyzes 5 competitors across 10 dimensions, flags disputed claims, and produces a comparison matrix with citations.

📚 Technical Documentation

Create presentations, documentation, or technical articles backed by solid research. Every claim is cited, confidence levels are tracked, and sources are evaluated for reliability.

Example: "Create a presentation on the state of AI coding assistants in 2026" → Research mode gathers 40+ notes from academic papers, product pages, and benchmarks, then Work mode builds a fully-cited presentation.

🐛 Root Cause Investigation

Research known issues, bug reports, and solutions for complex problems. Track which solutions are confirmed vs. speculative, and identify patterns across multiple sources.

Example: "Why is our WPF app freezing?" → Research mode finds 15 related issues, categorizes by root cause, tracks which solutions are confirmed, and produces a diagnostic report.

🎯 Best Practices Research

Research industry best practices, design patterns, and architectural approaches. Distinguish between confirmed best practices and speculative opinions.

Example: "What are the best practices for async/await in C#?" → Research mode gathers official guidance, community consensus, and known pitfalls, flagging areas where experts disagree.

Why Systematic Research Matters

❌ Without Research System

  • Shallow web search snippets
  • Ad-hoc note-taking in context window
  • Findings lost after conversation
  • No source attribution
  • No confidence tracking
  • Synthesis is guesswork
  • Can't identify gaps
  • No structured output

✅ With Research System

  • Deep reading of full sources
  • Structured notes with metadata
  • Persistent evidence accumulation
  • Full source attribution
  • Confidence levels tracked
  • Evidence-based synthesis
  • Systematic gap analysis
  • Professional research reports

Ready to Research Smarter?

Experience systematic, evidence-based research with Rekall's advanced Research system.