Your schedule has defects. Find them before your client does.
Upload a Primavera P6 export (.xer) and get a defensible DCMA 14-point health check in minutes — every finding cited to an activity ID, with a branded PDF you can hand straight to the project director.
Run your free health check
No signup, no credit card — just a work email so your report has a name on it.
Drag your .xer file here
or click to browse — up to 50 MB
Your file is processed to generate this report, retained for 30 days, then deleted. It is never used to train models.
The DCMA 14-point assessment
The industry-standard schedule quality checks, at their published thresholds — never loosened to make a schedule look better. Eight are free, forever.
Missing logic
Incomplete activities with no predecessor or successor. Threshold ≤ 5%.
FreeLeads (negative lag)
Negative lag hides work behind its predecessor. Zero tolerance.
FreeLags
Positive lag is invisible, unstatusable duration. Threshold ≤ 5%.
FreeRelationship types
At least 90% of links should be finish-to-start.
FreeHard constraints
Constrained dates override network logic. Threshold ≤ 5%.
FreeHigh float
Total float over 44 days usually means missing logic. Threshold ≤ 5%.
FreeNegative float
The plan cannot meet its committed dates. Zero tolerance.
FreeHigh duration
Remaining durations over 44 days are too coarse to status. Threshold ≤ 5%.
FreeInvalid dates
Actual dates after the data date, forecast dates before it.
PaidResources
Activities with duration but no resources assigned.
PaidMissed tasks
Activities finishing later than their baseline said.
PaidCritical path test
Inject a delay, re-schedule, and watch the finish date respond.
PaidCPLI
Critical path length index — how realistic the remaining plan is.
PaidBEI
Baseline execution index — done vs what the baseline required by now.
PaidHow it works
Export from P6
File → Export → XER. No add-ins, no macros, nothing installed on your machine, no access to your P6 database.
We parse it deterministically
Plain engineering code normalises your programme and runs every check at the published DCMA thresholds. Every number is computed; none is guessed.
Get a report you can send
A branded PDF with every claim cited to an activity ID, a relationship or a computed metric — ready for the project director.
Not another AI chatbot
In an industry that litigates, a confidently wrong number is worse than no number.
Deterministic core
All engineering — DCMA checks, float, critical path — is plain code. A language model never computes a number; it may only narrate figures the engine has already fixed.
Every claim cited
No sentence in any report without a traceable source: an activity ID, a relationship, or a computed metric. You can verify everything against your own programme.
Native Arabic
Built bilingual from day one, by engineers who work in the GCC — not machine-translated as an afterthought.
Your data stays yours
A contractor's schedule is commercially sensitive. Tenant isolation, 30-day retention, and your schedules are never used to train models.
Pricing
Start free. Upgrade when the report starts doing your Monday-morning arguing for you.
- All 14 checks
- Complete findings lists
- Excel export
- Everything in Individual
- Update-to-update comparison
- Power BI export
- Full Arabic reports
- Everything in Professional
- Shared workspaces
- Portfolio views
- SSO and audit logs
- On-premise deployment
- SLA and priority support
Frequently asked questions
Is my schedule data safe?
Your file is processed to generate the report, retained for 30 days, then deleted. Customer schedules are never used to train any model, and every customer's data is isolated.
Which file formats do you support?
Primavera P6 .xer exports. If your file cannot be parsed you get a message saying exactly why — and every real-world failure becomes a permanent regression test in our parser.
How is the score calculated?
Against the published DCMA 14-point thresholds, weighted by how badly each check is failed. Checks that cannot be assessed are excluded from the denominator — your schedule is never punished for data it could not supply — and thresholds are never loosened.
Why is the health check free?
The free report is our marketing. Eight structural checks are free forever; paying unlocks the remaining six, complete findings and update-to-update comparison.
Can the AI get the numbers wrong?
No. Every number is computed by deterministic engineering code. A language model never computes a number in Masar AI — that is an architecture rule, not a promise.
Do you really support Arabic?
Yes — the site and support are bilingual, and full Arabic reports ship with the Professional plan. Arabic is a foundation here, not a checkbox.