Skip to main content

Product Roadmap

Overview​

Cyzag Blueprint is intentionally phased to deliver value incrementally while building a solid foundation for future capabilities. Phase 1 delivers a focused, production-ready core: operator rounds and asset condition monitoring. Phases 2 and 3 layer on maintenance workflows without changing the underlying data model.

This phased approach allows teams to start capturing reliable asset data immediately, then progressively add intelligence (alerts, observations) and AI-powered insights on top of the same foundation.

Phase 1: Operator Rounds & Asset Readings​

Status: Current focus — the vision Blueprint is working towards

Phase 1 establishes the foundation for capturing asset condition data from the field.

Core Capabilities​

  • Assets, Asset Types, and Asset Attributes: Define equipment and the measurements that matter for each type
  • Measurement Points: Counters, temperature, pressure, vibration, and other readings tied to specific asset attributes
  • Offline-First Operator Rounds: Mobile apps (iOS and Android) guide operators through predefined checkpoints, capturing readings even without connectivity
  • Last Known Value & Historical Readings: Every reading updates the current state and contributes to historical trends for each asset attribute
  • Photos, Notes, and Observations: Operators capture free-form observations and photos at each checkpoint to supplement structured data
  • Background Sync: When connectivity returns, rounds and readings sync to the server automatically

Who It's For​

  • Operators: Walk the plant floor with a mobile device, complete rounds offline, capture structured readings quickly
  • Supervisors: Review completed rounds, check for abnormal readings, follow up on observations
  • Reliability Engineers: Analyze historical trends to identify degradation patterns and plan maintenance

What It Delivers​

A reliable system for capturing and storing asset condition data with strong offline support, designed for industrial environments where connectivity is unreliable.

Phase 2: Observations, Alerts & Triage​

Status: Future — builds on Phase 1 data model

Phase 2 adds intelligence to the data captured in Phase 1, turning abnormal readings and operator comments into structured observations and alerts.

Core Capabilities​

  • Thresholds and Rules: Define acceptable ranges for asset attributes (e.g., "discharge pressure should be 40–50 bar")
  • Automated Alerts: Flag readings that exceed thresholds or show unusual patterns
  • Observations with Severity: Promote free-form operator notes into structured observations with severity levels (low, medium, high, critical)
  • Alert Views: Dashboards for supervisors and engineers to see all abnormal conditions across the plant
  • Lightweight Triage: Acknowledge alerts, add comments, mark for follow-up, or query via AI agent for recommendations

Who It's For​

  • Supervisors: Get immediate visibility into abnormal conditions reported during rounds
  • Reliability Engineers: Set thresholds and rules to catch issues early
  • Shift Handovers: Ensure nothing important gets lost between shifts

What It Delivers​

A layer of intelligence that ensures abnormal readings and operator observations get noticed, triaged, and acted upon — without requiring operators to change how they work.

Phase 3: Agentic AI & MCP Server​

Status: Future — builds on Phase 2 observations and alerts

Phase 3 leverages AI agents and the Model Context Protocol (MCP) to enable natural language interaction with your maintenance data, intelligent recommendations, and context-aware insights.

Core Capabilities​

  • Natural Language Queries: Ask questions about asset readings, trends, and observations in plain English
  • AI Agents with Full Context: Agents understand your asset structure, historical trends, observations, and alerts
  • MCP Server: Exposes asset data and actions as tools that LLMs can securely query and invoke
  • Intelligent Recommendations: AI analyzes patterns to suggest maintenance priorities, identify root causes, and predict issues
  • Voice-Enabled: Mobile and web interfaces support voice queries for hands-free operation

Who It's For​

  • Supervisors: Quickly query asset status and critical issues using natural language
  • Reliability Engineers: Analyze trends and correlations across assets without writing SQL queries
  • Maintenance Teams: Get instant answers about asset history and recommended actions
  • Shift Handovers: Summarize key issues and abnormal conditions with a simple question

What It Delivers​

An intelligent layer that makes all your Phase 1 and Phase 2 data accessible through conversation, enabling faster insights and better decisions without navigating complex UIs or predefined workflows.

Design Philosophy​

Why This Sequence?​

  1. Phase 1 establishes the data foundation: assets, attributes, readings, and offline sync
  2. Phase 2 adds intelligence without changing how operators work or the underlying data model
  3. Phase 3 adds AI-powered insights and natural language access to all the data from Phases 1 and 2

Each phase delivers standalone value while preparing for the next.

What Stays Constant​

  • Data Model: The core entities (Assets, AssetTypes, AssetAttributes, Readings) remain stable across all phases
  • Offline-First: Mobile apps continue to work without connectivity
  • Platform Support: iOS, Android, and Web across all phases
  • Security Posture: Encryption at rest, secure transport, scoped access by environment

What Changes​

  • Intelligence: Phase 2 adds automated alerts and thresholds; Phase 3 adds AI-powered insights
  • Interaction Model: Phase 3 introduces natural language queries and agentic AI
  • Integrations: MCP servers enable LLM access to asset data; can connect to external tools as needed

Current State vs. Vision​

Blueprint is a reference implementation and lab, not a product. The documentation describes the vision we're building towards for each phase, even if not all features are fully implemented yet.

The goal is to demonstrate end-to-end patterns and best practices that can be reused in Cyzag projects, with Phase 1 as the current focus.

See Also​