Help & DocumentationAI Consulting Platform

AI Consulting Platform

Generate McKinsey-quality assessment reports using AI and consulting frameworks

Overview

The AI Consulting Platform transforms your document repository into actionable business intelligence. Using proven consulting frameworks (McKinsey 7S, Porter's Five Forces, SWOT, etc.) and advanced AI, it generates professional-grade assessment reports that would typically cost thousands from consulting firms.

Key Features

  • • Framework-based Analysis: Apply proven consulting frameworks to your documents
  • • Intelligent Mode: AI auto-detects query type and optimizes response strategy
  • • Multiple Output Formats: PDF, Word, Excel, or PowerPoint reports
  • • McKinsey-Quality Pipeline: Multi-step generation with SCQA and Pyramid Principle

Analysis Modes

The platform offers two distinct analysis modes to suit different needs:

DocAssure Analysis

Framework-based - Select a consulting framework and let AI analyze your documents against it.

  • • Choose from 10+ consulting frameworks
  • • Auto-discovers relevant documents
  • • Generates structured assessments
  • • Includes gap analysis and recommendations

Own Analysis

Custom selection - Manually select documents and ask questions without framework constraints.

  • • Manually select specific documents
  • • Ask any question about content
  • • Create custom analysis reports
  • • Maximum flexibility for ad-hoc research

Intelligent Mode

Intelligent Mode is a powerful feature that automatically detects what type of question you're asking and chooses the optimal strategy to answer it. This results in faster, more accurate responses.

Toggle Location

The Intelligent Mode toggle is located in the chat header, next to the "Generate Report" button. When enabled, it displays an amber highlight with a lightning bolt icon.

Intelligent Mode
← Enabled (recommended)

How It Works

  1. 1
    Intent Detection

    AI analyzes your question to classify it: metadata query, document retrieval, or analytical question.

  2. 2
    Strategy Selection

    Chooses between Database Query (fast SQL for statistics) or Document Search (RAG for content).

  3. 3
    Optimized Execution

    Runs the query using the optimal method - milliseconds for counts, seconds for content analysis.

  4. 4
    Transparent Feedback

    Shows you which strategy was used, the detected intent, and response time for transparency.

Query Types & Strategies

Query TypeStrategySpeedExample Questions
MetadataDatabase Query (SQL)10-100ms"How many documents?", "Storage usage?"
RetrievalDocument Search (RAG)1-5 seconds"Find compliance docs", "What does the policy say?"
AnalyticalSQL Analytics50-200ms"Documents by category", "Upload trends"

When to Disable Intelligent Mode

Keep Intelligent Mode enabled for most use cases. Only disable it if you specifically want to force all queries through the document search (RAG) system, regardless of query type. This might be useful when you want document evidence even for simple counting questions.

Consulting Frameworks

The platform supports industry-standard consulting frameworks for structured analysis:

Strategy Frameworks

  • • McKinsey 7S Model
  • • Porter's Five Forces
  • • SWOT Analysis
  • • PESTEL Analysis

Operational Frameworks

  • • Value Chain Analysis
  • • BCG Matrix
  • • Balanced Scorecard
  • • RACI Matrix

Risk & Compliance

  • • Risk Assessment Matrix
  • • Compliance Gap Analysis
  • • Control Effectiveness

Financial Analysis

  • • DuPont Analysis
  • • Cost-Benefit Analysis
  • • ROI Assessment

Output Formats

Reports can be generated in various formats to suit different presentation needs:

📝

Comprehensive (DOCX)

Full detailed report with all sections, evidence, and appendices

📄

Executive Summary (PDF)

Concise 2-3 page summary for leadership review

📊

Charts Focus (XLSX)

Data-rich spreadsheet with charts and visualizations

📽️

Presentation (PPTX)

Slide deck ready for stakeholder presentations

Generating Reports

Follow these steps to generate a consulting report:

  1. 1
    Select Analysis Mode

    Choose "DocAssure" for framework-based analysis or "Own Analysis" for custom queries.

  2. 2
    Choose Framework (DocAssure mode)

    Select the consulting framework that best fits your analysis needs.

  3. 3
    Select Output Format

    Choose the document format (PDF, DOCX, XLSX, or PPTX).

  4. 4
    Select Documents

    Use Auto mode for AI-recommended documents or Manual to select specific files.

  5. 5
    Click Generate Report

    The AI will analyze your documents and generate the report. This may take 15-60 seconds.

McKinsey-Quality Pipeline

For DocAssure mode, the system uses a 5-step pipeline: Deep Analysis → Pyramid Structure → Executive Summary (SCQA) → Action Titles → Strategic Recommendations. Progress is shown in real-time.

Reports Panel

Generated reports are saved to the Reports Panel on the left sidebar. From there you can:

  • View reports - Click a report card to open the full document viewer
  • Download reports - Click the download icon to save to your computer
  • Delete reports - Click the trash icon to remove (this cannot be undone)
  • Track confidence - Each report shows a confidence score based on document coverage

Report Persistence

Reports are automatically saved to your browser's local storage. They will persist across page refreshes but are specific to your browser. For permanent storage, download the report file.

Best Practices

Do

  • ✓Select documents relevant to your framework
  • ✓Keep Intelligent Mode enabled for faster responses
  • ✓Use the chat to ask follow-up questions about reports
  • ✓Download important reports for permanent storage

Avoid

  • ✗Selecting too few documents (aim for 3-10)
  • ✗Mixing unrelated document types
  • ✗Generating reports without reviewing the framework requirements
  • ✗Relying solely on browser storage for critical reports

Troubleshooting

Report not appearing in Reports Panel

Possible causes:

  • Generation failed (check for error message in chat)
  • Browser storage is full or disabled
  • Page was refreshed before generation completed

Solution: Check browser console (F12) for errors. Try generating again.

Low confidence score

Possible causes:

  • Selected documents don't cover all framework requirements
  • Documents are too short or lack detailed content
  • Document types don't match framework expectations

Solution: Add more relevant documents. Check the "Missing Elements" section in the report.

Generation takes too long

Possible causes:

  • Too many documents selected (more than 20)
  • Very large documents being analyzed
  • AI model under heavy load

Solution: Reduce document selection to 5-10 most relevant files. Try during off-peak hours.