What Qlira AI does, and what it leaves to you.
The AI keeps the plan current: it replans, reschedules and assigns after every update, reads reports, answers questions and drafts report-outs. The forecast is something you can check, every risk flag shows its evidence, and no plan changes without a person accepting.
- 01Runs after every update
Plans, replans, schedules and assigns
After every update it recomputes the critical path, forecasts the finish, reschedules the work ahead, scores the actions that recover time and proposes tomorrow’s plan for each crew. People accept, adjust or dismiss.
- 02Reads with Claude
Reads reports and minutes
Progress reports, meeting minutes and forwarded updates, whatever the format, become dated progress updates and issues mapped to tasks. You review the extraction before any of it is applied.
- 03Only from what you may see
Answers with citations
Why is this task late? What is waiting on the client? The answer cites the task, report, issue or assessment it comes from, drawn only from what the asker may see.
- 04You edit before it sends
Drafts report-outs
Weekly, monthly and HSE reports drafted from the KPIs and the log. A person edits and sends; nothing leaves on its own.
- 05From your projects only
Learns how long your work really takes
From your own finished activities, per company. It is only used once it forecasts better than the plan did, and it keeps learning after each import.
- 06Checkable, on purpose
Forecasts and flags, and shows its work
The critical path, the finish simulation and the thirteen risk checks give the same answer for the same inputs, and every flag carries its evidence.
Why is STR-1188 late?
Grounded in this project’s schedule, reports, issues and event log. Every claim cites its source.
Level 11 columns is 6 working days behind its baseline finish of 9 Feb. Work on gridlines C to E stopped on 10 Feb because rebar cannot be fixed until shop drawing SD-042 rev C is approved [3]; the drawing has been with the consultant for 11 days [3].
The task sits on the critical path with no float [4], so each day of hold moves the Level 12 slab and the structure milestone by a day. The supervisor’s last report shows 78% complete with 6 crew on the remaining gridlines [2].
Escalating the approval is the recommended action; the what-if gains 1.0 day on the finish for each day the hold is shortened [4].
Confidence high
Answers with sources
Ask about any task, delay or decision. The answer cites what it used and says how confident it is.
Assessing the project
The engine rebuilds the forecast from the live schedule and every report since the last run.
- Snapshot1,093 tasks, their links, calendars, daily reports, permits and open deliveries as of today.
- Work calendarsWorking days, hours and holidays per task, so durations count site time.
- Critical pathForward and backward pass over FS / SS / FF / SF links: early and late dates, float, the critical chain.
- Learned durationsHow long this company's activities really take, from its finished work; reported pace overrides it.
- Finish simulation300 runs of the remaining work: the likely finish, its range and the chance of a late one.
- Risk checks13 checks: pace, float, readiness, blockers, stale reports, permits, supply, forecast. Each with its evidence.
- RecommendationsCandidate actions scored by their what-if gain on the finish date.
- Write-backDates, float and criticality onto every task, and an event on the log.
Done in 14.8 s. 7 risks flagged, 4 recommendations, forecast +6.5 d against the plan.
Replans in the open
Every step of an assessment is visible while it runs, and the result lands on the last line.
Key functionality by module.
Every change in Qlira is an event on the project’s log, so every module below shares one history and one set of permissions.
The live plan
One schedule everyone works from, with the critical path and the forecast finish on the same bars as what the field reported.
Activities
1,093
Matched
1,041
New
43
Needs a look
9
| In the file | Task |
|---|---|
| A1240L12 SLAB POUR ZONE A+B | STR-1240 |
| A1188L11 COLUMNS GL A-F | STR-1188 |
| A1305L13 COLUMNS (REV) | STR-1301 |
| A2117CHW RISERS L10-14 DELIVERY | New task |
| A0742CABLE CONTAINMENT L9 | MEP-0742 |
Apply as baseline “BL R3 · 13 Feb”. Progress the site has reported is never overwritten.
Imports with a mapping review
Activities matched by id, code, alias or name. You see only the rows that need a decision.
Open risks
2 high · 3 medium
Level 11 columns is 6 days behind and drives the finish
“Rebar for gridlines C to E held pending shop drawing approval.”
PO-0417 risers due 21 Feb; supplier delivered 62% of orders on time in 90 days
Total float down to 1.5 days from 9
Last daily report for Blockwork L8 was 3 days ago
Hot-work permit for cable containment L9 expires tomorrow
Risks with evidence
Thirteen checks covering pace, float, readiness, blockers, stale updates, permits, supply and the forecast. Each flag shows what it saw.
On the horizon
What starts and finishes next, on the dates the engine forecasts.
Next 3 days
14 – 17 Feb
1 starting · 2 finishing · 1 critical · 1 behind
Next 3 weeks
17 Feb – 7 Mar
3 starting · 2 finishing · 2 critical
Next 3 months
7 Mar – 16 May
14 starting · 11 finishing · 2 milestones · 4 critical
Three days, three weeks, three months
Starts, finishes and milestones on forecast dates, with critical and behind marked.
S-curve
Cumulative progress, weighted by planned hours. Baseline BL R2.
- Baseline
- Current plan
- Actual
- Forecast
KPIs and the S-curve
Progress, schedule performance, days late and milestones with trends; baseline, plan, actual and forecast on one curve.
- Schedule
- Import from Primavera, Microsoft Project, Excel, PDF and more, with mapping review
- Baselines; what the field reported is never overwritten
- Gantt: baseline, plan, progress, critical path, forecast marks
- Dependency editing (FS / SS / FF / SF), task pages with history
- Daily loop
- Daily updates per task from the phone, with photos and remarks
- Engineer review with cause tags
- Next-day plans: issued by the AI, adjusted with a reason, approved
- Issues and escalations routed up the reporting line
- Offline saves that replay when the signal returns
- Insights
- Critical path with each task’s own working calendar
- Likely finish date, its range and the chance of a late finish
- 13 risk checks, each with its evidence
- Ranked recommendations with the days each recovers
- Automatic re-assessment after every change
- KPIs
- Progress, schedule performance, days late, milestones
- Field, quality and money groups with trends
- Team table by supervisor and engineer
- S-curve: baseline, plan, actual, forecast
- Cash-flow curve
- Notifications
- Derived from the event log, never from a button
- In-app, email and web push
- Per-person preferences and muted categories
- Safety
- Permits requested, verified, rejected; missing permits flagged by the engine
- Daily duties checklist, observations, incidents with investigations
- Training register, HSE daily / weekly / monthly report-outs
- Procurement
- Suppliers with 90-day reliability
- Requisitions, quotes, award, purchase orders in five states
- Deliveries; supply risk on the critical path flagged by the engine
- Accounting
- Cost lines and ledger with after-close audit
- Applications and variations, petty cash
- Month close with a checklist; time and cost verdict on the KPIs
- Deployment
- Hosted, or on your own servers
- Roles are data: rename and re-permission any of the ten templates
- Single sign-on, audit export
Start with one project.
Tell us about your projects and the tools you plan in. We set your team up, bring in a schedule and walk you through the first assessment together.