AI Copilot
Applies to: Free · Pro · Enterprise Last updated: 08-Jun-2026
Overview
The Kazinex AI Copilot is a conversational assistant that understands your schedule data. Ask questions in plain English, get instant insights, and let the AI make targeted edits — all without leaving the application. The Copilot can query activities, analyse float distribution, check quality, generate reports, export professional DOCX documents, manage grid columns, and even create entire project schedules from industry templates.
Opening the Copilot
There are two ways to open the AI Copilot panel:
- Toolbar button — click the AI Copilot button (sparkles icon) in the main toolbar.
- Keyboard shortcut — press Ctrl + Shift + I to toggle the panel. Press Escape to close it.
The Copilot requires a schedule to be loaded. The button is disabled when no schedule is open.
Panel Layout
The AI Copilot appears as a resizable side panel on the right edge of the screen:
| Property | Default | Range |
|---|---|---|
| Width | 420 px | 320 – 700 px |
| Animation | 0.3 s slide | — |
The panel contains:
- Header — model selector, new conversation button, close button
- Message list — scrollable conversation with AI responses, tool results, and confirmation cards
- Input bar — text field for your questions and instructions
AI Models
Kazinex offers five AI models with different capabilities and tier requirements:
| Model | Provider | Tier | Cost | Best For |
|---|---|---|---|---|
| GPT-4.1 Mini | OpenAI | Free, Pro, Enterprise | $ | Most queries — fast and efficient |
| DeepSeek V3 | DeepSeek | Free, Pro, Enterprise | $ | Strong reasoning and tool use |
| GPT-4.1 | OpenAI | Pro, Enterprise | $$ | Advanced reasoning for complex analysis |
| GPT-4o | OpenAI | Pro, Enterprise | $$$ | Multi-modal flagship model |
| Claude Sonnet 4 | Anthropic | Pro, Enterprise | $$$$ | Excellent at structured output |
Switch models using the model selector dropdown in the panel header. Locked models show a padlock icon with a "PRO" label — upgrade your plan to unlock them.
Your model selection is remembered across sessions.
What the Copilot Can Do
The Copilot has access to 79 tools across 6 categories that interact with your schedule data:
| Category | Tools | Description |
|---|---|---|
| Read Tools | 15 | Query, search, and analyse schedule data |
| Write Tools | 34 | Edit activities, WBS, relationships, resources, calendars, codes |
| Integration Tools | 18 | Quality, comparison, grid filtering, narratives, navigation |
| App Feature Tools | 12 | Grid columns, presets, grouping, sorting, export, theme |
| Export Tools | 3 | DOCX generation, narrative export, download |
| Knowledge Tools | 8 | Industry standards, delay analysis, EVM, P6 guidance |
See the dedicated pages for details:
- Asking Questions — querying, analysis, and insight tools (15 read tools)
- AI Modifications — editing, bulk updates, and structural changes (34 write tools)
- Grid & App Features — column management, presets, grouping, sorting, export, theme (12 tools)
- Narratives & Export — report generation, DOCX export, professional documents (6 tools)
- Knowledge & Standards — industry knowledge, delay analysis, EVM, best practices (8 tools)
- Baseline & Comparison — baseline awareness, variance analysis, comparison workflow
- Project Generation — creating new schedules from templates
- Confirmation Workflow — reviewing and approving proposed changes
Welcome Screen Suggestions
When you open the Copilot with no conversation history, you'll see six quick-start suggestions:
- Negative Float — "Show me activities with negative float"
- Quality Score — "What's the quality score of this schedule?"
- Critical Path — "Show activities on the critical path"
- Schedule Statistics — "Give me an overview of this schedule"
- Update Activity — "Update the duration of [activity]"
- Paste Progress — "Paste progress data from Excel"
An AI Wizard button is also available for guided project generation.
Context Awareness
The Copilot injects live schedule context into every request, including:
- Project details — name, ID, data date, activity/WBS/resource counts
- Baseline status — whether a baseline is loaded, baseline project name, data date, and completion
- WBS hierarchy — the full work breakdown structure tree
- Calendars — all calendars with working patterns and activity counts
- Activity codes — available code types
- UDF types — available user-defined fields
This context ensures the AI understands your specific project and can provide accurate, targeted responses.
Baseline Awareness
When a baseline schedule is loaded, the Copilot automatically:
- Shows current vs baseline values side by side
- Calculates and highlights variance for dates, duration, float, and percentages
- Flags activities that slipped, improved, were added, or deleted
- Offers to add BL and Variance columns when showing progress data
See Baseline & Comparison for full details.
Smart Column Matching
The Copilot uses fuzzy name matching to understand column requests in natural language:
- British spelling — "labour" matches "labor" columns
- Abbreviations — "pct", "percent", "%" all match percentage columns
- Word separation — "non labour" matches "nonlabor" columns
- Partial names — "total float" matches "Total Float Hr Count"
For example, saying "add non labour percent to grid" correctly adds the Nonlabor Units % Complete column.
Token Limits
AI usage is metered by tokens (a unit of text processed by the model):
| Tier | Monthly Tokens | Daily Requests |
|---|---|---|
| Free | 50,000 | 25 |
| Pro | 500,000 | 500 |
| Enterprise | Unlimited | Unlimited |
Usage warnings appear at 80% and 90% of your limits. When limits are reached, you'll see a notification suggesting an upgrade.
Token counters reset on the 1st of each month (monthly) and at midnight (daily).
Conversations
Each conversation is automatically saved and can be resumed later:
- Auto-title — the conversation title is generated from your first message (max 80 characters).
- Per-project — conversations are associated with the loaded schedule.
- Local-first — stored in your browser's localStorage with optional background sync to the cloud.
- Maximum 50 conversations — oldest conversations are trimmed automatically.
Start a new conversation by clicking the refresh icon in the panel header.
Complete Tool Reference
Below is the full list of all 79 tools the Copilot can use. Click each category link for detailed documentation, parameters, and example prompts.
Read Tools (15) → Asking Questions
| Tool | Description |
|---|---|
query_activities | Search and filter activities by status, WBS, text, float, dates, codes |
get_activity_details | Full details for a single activity |
get_schedule_statistics | Aggregate schedule statistics |
get_critical_path | Critical and near-critical path activities |
get_quality_score | Schedule quality health checks |
search_schedule | Free-text search across all entities |
get_wbs_tree | WBS hierarchy tree |
get_progress_summary | Overall progress overview |
get_float_analysis | Float distribution analysis |
get_resource_usage | Resource and assignment aggregation |
get_milestone_status | Milestone status summary |
get_activities_by_code | Filter by activity code type/value |
get_date_analysis | Project date window and delay counts |
get_cost_summary | Budget vs actual cost analysis |
get_calendars | Calendar list with working patterns |
Write Tools (34) → AI Modifications
| Tool | Description |
|---|---|
update_activity | Update a single field on one activity |
add_activity | Create a new activity under a WBS |
delete_activity | Remove an activity and its relationships |
move_activity | Move activity to a different WBS |
duplicate_activity | Deep copy an activity |
dissolve_activity | Remove and rewire predecessors to successors |
add_relationship | Create a dependency (FS/SS/FF/SF) |
delete_relationship | Remove a dependency |
update_relationship | Change type or lag of a relationship |
add_wbs | Create a new WBS node |
rename_wbs | Rename a WBS node |
delete_wbs | Remove an empty WBS node |
move_wbs | Move a WBS node to a new parent |
add_resource | Create a new resource |
delete_resource | Remove a resource from the pool |
assign_resource | Assign a resource to an activity |
delete_assignment | Remove a resource assignment |
add_activity_code_type | Create a new code type |
add_activity_code_value | Add a value under a code type |
assign_activity_code | Assign a code value to an activity |
unassign_activity_code | Remove a code assignment |
set_udf_value | Set a user-defined field value |
add_calendar | Create a new calendar |
update_calendar | Modify calendar properties |
add_calendar_exception | Add holidays/non-working days |
delete_calendar | Remove a calendar |
update_project | Change project-level settings |
bulk_update_activities | Update fields across multiple activities |
bulk_transform | Pattern-based bulk transformation |
apply_progress | Apply actual dates and % complete |
make_at_completion_equal_budget | Align at-completion with budget |
recalculate_schedule | Run CPM forward/backward pass |
plan_project_generation | Generate a project from templates |
import_schedule | Import detailed schedule data |
Integration Tools (18) → Baseline & Comparison
| Tool | Description |
|---|---|
get_quality_results | Latest quality check results |
get_quality_issues | Quality issues by check/category/severity |
get_quality_resolution | Fix guidance for a quality check |
get_comparison_results | Baseline comparison summary |
get_comparison_changes | Comparison changes by type/impact |
run_quality_check | Trigger quality check and return results |
run_comparison | Trigger/check baseline comparison |
filter_activities_in_grid | Apply a filter to the grid |
filter_quality_issues_in_grid | Filter grid by quality issue |
filter_comparison_in_grid | Filter grid by comparison change |
clear_grid_filter | Remove active grid filter |
navigate_to_tab | Switch to a specific UI tab |
generate_schedule_narrative | Full programme narrative |
generate_quality_narrative | Quality assessment narrative |
generate_comparison_narrative | Comparison variance narrative |
get_productivity_rates | Construction productivity rates |
get_schedule_best_practices | Scheduling best practice guidance |
export_narrative | Export narrative as DOCX/Markdown |
App Feature Tools (12) → Grid & App Features
| Tool | Description |
|---|---|
list_available_columns | List all 150+ grid columns |
add_columns_to_grid | Add columns by name (fuzzy matched) |
remove_columns_from_grid | Remove columns by name |
set_column_visibility | Show/hide specific columns |
list_column_presets | List available column presets |
apply_column_preset | Apply a named column preset |
set_grouping | Set grid grouping mode |
expand_collapse_groups | Expand or collapse all groups |
sort_grid | Sort grid by a column |
clear_sort | Remove all sorting |
export_data | Export schedule (XER/XML/MPP) |
set_theme | Switch light/dark/auto theme |
Knowledge Tools (8) → Knowledge & Standards
| Tool | Description |
|---|---|
get_delay_analysis_guidance | SCL Protocol / AACE delay analysis methods |
get_evm_metrics | Earned Value and Earned Schedule calculations |
get_p6_scheduling_guidance | P6 scheduling options and types |
get_contract_schedule_requirements | NEC4/FIDIC/JCT contract requirements |
get_reporting_template | Report structure templates |
get_s_curve_data | Early/late/actual S-curve data points |
get_productivity_rates | Industry productivity rates by trade |
get_schedule_best_practices | PMI/DCMA/AACE best practice guidance |
Tips & Best Practices
- Be specific — "Show activities in WBS 1.2 with float under 5 days" gets better results than "show me some activities".
- Use the right model — GPT-4.1 Mini handles most queries perfectly. Switch to GPT-4.1 or Claude for complex multi-step analysis.
- Review before applying — all write operations show a confirmation card. Take a moment to read the summary before clicking Apply.
- Chain questions — the Copilot remembers your conversation context. Follow up with "now filter those to only critical path" instead of repeating the full query.
- Ask for everything — when asking about completion %, the Copilot shows ALL percentage types (labour units, non-labour units, cost, schedule, performance) by default.
- Leverage baseline — when a baseline is loaded, the Copilot automatically shows variance data. Ask "what changed from baseline?" for a complete analysis.
- Use natural language for columns — say "add non labour percent to grid" and the fuzzy matcher handles spelling and abbreviation differences.