Skip to main content

Max AI Guide

Max AI is MachineMetrics' agentic intelligence layer — a team of role-based digital agents built to interpret, adapt, and execute production tasks every day. Max connects real-time data from machines, ERP systems, and tribal knowledge to deliver foresight and guidance that aligns teams, preserves expertise, and drives performance.

1. What is Max AI?​

Max AI addresses the execution gap — the disconnect between production plans and shop floor reality. While plans assume smooth operations, the reality includes machine breakdowns, long setups, late materials, and tribal knowledge that overrides system logic.

Max AI bridges this gap by:

  • Connecting real-time data from machines, ERP, and documentation
  • Unifying data into a structured model AI can understand
  • Acting through role-based agents that guide and automate work

Key Capabilities​

CapabilityDescription
Chat with Your DataAsk questions in natural language and get instant answers from real-time data, ERP records, and tribal knowledge
Instant VisualsGenerate reports, charts, and downtime breakdowns on demand — no manual building required
Agentic WorkflowsAutomate routine decisions like flagging late jobs, rescheduling work, and updating ERP cycle times

2. Accessing Max AI​

From Any Page​

  1. Click the Max AI icon in the navigation bar (typically bottom-right or top-right)
  2. The Max AI chat panel opens
  3. Type your question or request

Keyboard Shortcut​

  • Press Ctrl + K (Windows) or Cmd + K (Mac) to open Max AI from anywhere in the platform

3. AI Agents​

Max AI provides specialized agents for different roles and tasks. Each agent is optimized for specific workflows.

3.1 Continuous Improvement Agent (Available)​

Your CI team that never sleeps — powered by Lean/Six Sigma methodologies.

Capabilities:

FeatureDescription
Surface BottlenecksIdentify production bottlenecks in real-time
Setup AnalysisAnalyze setup time trends and variations
Delivery PredictionMonitor delivery performance and predict risks
Shift ComparisonsGenerate shift performance comparisons

Methodologies Used:

  • Pareto Analysis — Identify the vital few causes of problems
  • 5 Whys — Root cause analysis for recurring issues
  • SMED — Single-Minute Exchange of Die for setup reduction
  • SPC — Statistical Process Control for quality monitoring

Example Prompts:

"What are the top 5 causes of downtime this week?"
"Show me setup time trends for CNC Mill 3 over the last month"
"Which jobs are at risk of being late?"
"Compare shift performance for the past 7 days"
"What's causing the longest cycle times on operation 12345?"

3.2 Operator Assistance Agent (Coming Soon)​

AI-driven guidance for the frontline — helping operators start every shift ready.

Planned Capabilities:

FeatureDescription
Auto Handover BriefsStart each shift with status, jobs, and priority issues
Instant Doc AccessSetup instructions, work orders, and specs surface by job phase
AI TroubleshootingGet help from Knowledge Hub content and historical feedback

3.3 Production Management Agent (Coming Soon)​

Generate insights, not more data — for supervisors and production managers.

Planned Capabilities:

FeatureDescription
Tribal KnowledgeCapture and surface SOPs, notes, and expertise when it matters
Proactive AlertsFlag risks, delays, and quality concerns in real time
Production VisibilityMonitor jobs, machines, and schedule health across the floor

3.4 Shop Floor Scheduling Agent (Coming Soon)​

Adjust the schedule in real-time — keeping production on track.

Planned Capabilities:

FeatureDescription
At-Risk Job FlaggingIdentify jobs before they're late
Auto ERP UpdatesUpdate cycle times in ERP from actual production data
Schedule HealthSurface schedule health across your floor
Capacity ExposureReveal hidden capacity across all resources

3.5 Onboarding Agent (Coming Soon)​

Get up and running faster — for new users and new implementations.

Planned Capabilities:

FeatureDescription
AI-Guided SetupConnect machines and ERP with step-by-step contextual guidance
Role-Based WalkthroughsOnboarding adapts to your role, showing only what's relevant
Smart ConfigurationRecommendations for best dashboards and workflows

4. Chat with Your Data​

Max AI understands both the question and the context. Whether you're on the floor or in a meeting, get instant answers pulled from real-time data, ERP records, and tribal knowledge.

How It Works​

  1. Ask a Question — Type in natural language
  2. Max Interprets — Understands intent and context
  3. Data Retrieved — Pulls from machines, ERP, Knowledge Hub
  4. Answer Delivered — Clear response with supporting data

Question Types​

Production Questions:

"What's the current utilization across all machines?"
"How many parts did we make yesterday?"
"Which machines are currently down?"
"What's the OEE for Mill 5 this week?"

Analysis Questions:

"Why did OEE drop on Tuesday?"
"What's causing scrap on operation 7890?"
"Show me the trend of downtime for the past 30 days"
"Compare performance between first and second shift"

Guidance Questions:

"How do I set up a new workflow automation?"
"What's the best way to configure cycle time goals?"
"How do I add a widget to my dashboard?"

Tips for Better Results​

DoDon't
Be specific about time rangesUse vague timeframes
Name specific machines or operationsAssume Max knows what you're looking at
Ask one question at a timeCombine multiple unrelated questions
Include context (shift, machine group)Leave out relevant filters

Good Example:

"What were the top 3 downtime reasons for CNC Cell A during first shift last week?"

Less Effective:

"Why is downtime bad?"

5. Instant Visuals​

Need a shift summary or downtime breakdown? Just ask. Max AI builds reports in seconds, complete with visuals, trends, and root cause insights.

Generating Charts and Reports​

Example Requests:

"Create a downtime Pareto for this week"
"Show me a utilization trend chart for the past month"
"Generate a shift comparison report for yesterday"
"Build an OEE breakdown chart for Mill 3"

Chart Types Available​

Chart TypeBest For
ParetoIdentifying top contributors (downtime, rejects)
Trend LineShowing changes over time
Bar ChartComparing categories or groups
Pie/DonutShowing proportions
TableDetailed data with multiple columns

Adding Charts to Dashboards​

  1. Generate a chart using Max AI
  2. Click Add to Dashboard on the generated visual
  3. Select the target dashboard
  4. Choose position and size
  5. Click Save

The chart becomes a live widget that updates automatically.

6. Agentic Workflows​

Max AI handles routine decisions automatically — giving your team time back to focus on what matters most.

Automated Actions​

ActionDescription
Flag Late JobsAutomatically identify jobs at risk of missing delivery
Reschedule WorkSuggest or execute schedule adjustments based on real-time status
Update ERPPush actual cycle times back to ERP systems
Send AlertsTrigger notifications when thresholds are exceeded

Configuring Agentic Workflows​

  1. Go to Automations in the main menu
  2. Click Create Workflow
  3. Select a trigger (e.g., "Job at risk of being late")
  4. Define the action (e.g., "Notify supervisor", "Update priority")
  5. Set conditions and filters
  6. Enable the workflow

Example Workflow: Late Job Alert​

Trigger: Job estimated completion exceeds due date Condition: Job priority is "High" or "Critical" Action: Send notification to Production Manager with job details and recommended actions

7. Data Sources​

Max AI gets its intelligence from three key data sources:

7.1 Machine Data​

Real-time performance from any equipment — old, new, CNC, or manual.

  • Cycle times and part counts
  • Machine states (running, idle, down)
  • Alarms and conditions
  • Spindle loads and feed rates

7.2 ERP Integration​

Work orders, routings, and schedules sync directly with ERP.

  • Job information and due dates
  • Planned vs. actual quantities
  • Routing and operation details
  • Customer and order context

7.3 Knowledge Hub​

Tribal knowledge captured and made searchable.

  • SOPs and work instructions
  • Setup procedures
  • Troubleshooting guides
  • Best practices and notes

See the Knowledge Hub Guide for details on uploading and managing documentation.

8. Common Use Cases​

Daily Shift Handover​

"Give me a shift handover summary for second shift"

Max provides:

  • Jobs completed vs. planned
  • Current machine status
  • Open issues and alerts
  • Priority items for next shift

Root Cause Analysis​

"Why did utilization drop on Thursday?"

Max analyzes:

  • Downtime events and durations
  • Alarm patterns
  • Setup time anomalies
  • Provides ranked contributing factors

Performance Tracking​

"How are we tracking against our monthly OEE goal?"

Max shows:

  • Current OEE vs. target
  • Trend over the month
  • Gap analysis
  • Recommendations for improvement

Troubleshooting​

"Machine 7 keeps throwing alarm 1234, what should I check?"

Max searches Knowledge Hub for:

  • Previous occurrences
  • Resolution steps
  • Related documentation
  • Expert recommendations

9. Best Practices​

Getting Started​

  1. Start with questions you already know the answer to — Verify Max's accuracy
  2. Explore different question types — Production, analysis, guidance
  3. Use specific context — Machine names, date ranges, shifts
  4. Review and validate — Max is a tool to augment, not replace, expertise

Building Trust​

  • Compare Max's answers against reports you've built manually
  • Use Max for quick lookups, then dive deeper when needed
  • Provide feedback when answers aren't helpful

Team Adoption​

  • Share useful prompts with colleagues
  • Create a "prompt library" for common questions
  • Train operators on basic Max AI usage
  • Encourage feedback to improve results

10. Troubleshooting​

Max Doesn't Understand My Question​

  • Rephrase using simpler language
  • Break complex questions into smaller parts
  • Include specific machine or operation names
  • Add time context (today, this week, last month)

Data Seems Incorrect​

  • Verify the time range Max is using
  • Check if filters are applied correctly
  • Confirm machine names match exactly
  • Review data mapping in Machine Settings

Charts Won't Generate​

  • Ensure you have data for the requested time period
  • Check permissions for the machines/operations requested
  • Try a simpler chart type first

Knowledge Hub Content Not Found​

  • Verify documents are uploaded and indexed
  • Check document tags and categories
  • Use exact terminology from uploaded documents

11. Privacy and Security​

Data Handling​

  • Max AI runs on AWS Bedrock within MachineMetrics' private AWS tenant
  • Customer data is never used for training AI models
  • All queries are processed in isolated environments
  • Data remains within your MachineMetrics account

Access Control​

  • Max AI respects existing role-based permissions
  • Users only see data they have access to
  • Audit logs track AI interactions

Getting Help​

  • In-App: Click the help icon or ask Max AI "How do I get support?"
  • Support Portal: Visit the MachineMetrics Support Portal
  • Documentation: Browse the full documentation library