docReplay
Large-Scale Retrospective Intelligence: Deep analysis of historical performance data
Replay is a powerful Rust-based desktop application designed for deep, retrospective analysis of existing documents and historical performance data across vast collections. It allows users to analyze reports, transcripts, logs, and other text-based files against defined operational frameworks (Blueprints).
Core Capabilities
High-throughput analysis across multiple files. Process large document collections efficiently with configurable thresholds and parallel processing capabilities.
- Multi-file batch processing
- Configurable processing thresholds
- Efficient resource utilization
Long-term trend analysis and catalyst identification. Discover patterns across historical data to understand operational evolution and identify excellence.
- Trend analysis over time
- Catalyst identification
- Pattern evolution tracking
Automated justification generation for findings. Every score and insight is backed by specific evidence from your documents, ensuring transparency and trust.
- Automated rationale generation
- Evidence linking to source
- Transparent scoring
Efficient processing for large datasets. Built in Rust for maximum performance, with configurable thresholds and optimization options.
- Rust-based performance
- Multi-threaded processing
- Optimized resource usage
Typical Workflow
Launch Replay, log in, and select or create a Blueprint that defines your analysis framework. The Blueprint contains the Protocols and Metrics that will guide your analysis.
Load documents into your project via drag-and-drop. Replay estimates analysis time and cost (Insight Cost) before processing, ensuring transparency.
Optionally enable Entity Mapping to create an editable Mind Map of entities and relationships. This uses Consensus AI (EVA) and incurs Generative Cost.
Run core analysis using IBF. Scores content against Protocols, generates evidence (Markers), and calculates Metric/OEI scores. Optionally generate rationale for deeper insights.
Interact with the State of Entities view: Mind Map, Heatmap, Radar Chart, Timeline, and Evidence Table. Export data and generate custom reports as needed.
Replay presents analysis results through a multi-faceted interface that provides different perspectives on your data:
Mind Map
Visualizes entities and relationships discovered in your documents
Heatmap
Shows trends and concentrations of scores across topics
Radar Chart
Displays performance across key Protocols and Metrics
Timeline
Chronological sequence of key events and findings
Evidence Table
Detailed, sortable, filterable table with Protocol/Metric scores, Markers, Relationships, Sources, and optional Rationale
Cost Model
Replay uses a transparent cost model based on actual resource consumption, measured in Interaction Compute Seconds (ICS):
Insight Cost
Applied during Feature Extraction. Based on computational time for local IBF analysis.
Charged per unit of time, capped at the estimate provided before analysis.
Generative Cost
Applied when using Consensus AI features (Entity Mapping, Rationale, Reports).
Based on amount of text generated or tokens consumed.
Free Operations
Initial file extraction (no mapping), Blueprint management, and exploring results are free.
Licensing Models
- •Primarily PDF files (OCR and native)
- •Single analysis thread per organization
- •Restricted generative features
- •Limited processing speed
- Unrestricted file type support
- Multiple concurrent processing threads
- Full Entity Mapping and Rationale
- Optimized multi-threaded processing
Application Type
Standalone desktop application built in Rust. Runs locally on your machine.
Supported Platforms
Windows 10/11, macOS (recent versions), Linux (major distributions)
Minimum Requirements
Multicore processor, 16GB RAM minimum, SSD storage recommended
Privacy-First
Core analysis runs locally. Optional generative features send only necessary context.