Autonomous agents

Your project's AI team. They propose — you approve.

Eleven specialist agents on one runtime. Each does one job well, every number comes from a deterministic engine (CPM, leveling, DCMA, Monte Carlo) — and nothing touches your schedule or your subcontractors without your sign-off.

Your always-on crew

Two agents that run your week

Turn them on per project. They work on a cadence and draft into your inbox — and never send or change anything on their own.

Progress Collection
WEEKLY · YOU APPROVE
Chases each subcontractor as their work comes due.
1
Drafts the chase to each sub
2
Sub replies via a no-login link
3
You approve the updates
“Chase the trades for % complete and forecast dates”
Weekly Report
WEEKLY · YOU APPROVE
Drafts your client report on the day you choose.
1
Pulls the week's progress
2
Flags stale / inconsistent updates
3
You review & share
“Draft this week's progress report for the client”
On demand, in chat

The rest of the fleet

Ask in plain English and the right agent answers, drafts, or simulates — then hands you a proposal to approve, edit, or reject.

Schedule Generation
A brief becomes a full draft programme — WBS, logic, resources.
“16-week house renovation, 3 trades”
Proposes — you approve
Schedule Analysis
Charts, S-curves, EVM and health, on demand.
“Show the progress S-curve”
Read-only
Schedule Diagnostics
A DCMA 14-point quality check with the offending tasks.
“Run a DCMA check — what fails?”
Read-only
Schedule Risk
Monte-Carlo: P10–P90 finish dates and the odds you hit your date.
“How likely are we to finish on time?”
Read-only
Forecast
Projects your finish from SPI; cost (EAC) when actual costs are loaded.
“Where will we finish at this rate?”
Read-only
What-If Scenario
Simulate a change and see the impact before committing it.
“What if structure slips 10 days?”
Read-only
Delay Investigation
Finds the driver against baseline and drafts recovery options.
“Why are we behind — how do we recover?”
Proposes — you approve
Claims & Forensic
A claim-ready EOT/delay narrative grounded in your baseline — gaps flagged.
“Draft an EOT narrative for the delay”
Read-only
Resource Levelling
Within-float moves that smooth labour peaks — your finish holds.
“Where am I over-allocated?”
Proposes — you approve
Agents propose. Humans approve. Every number comes from a deterministic engine — nothing is written on its own.

See the agents on your own schedule.

Upload a P6 or MSP file, or generate one in chat — and put the fleet to work in minutes.

Get started free →