Plan a new project
Pick the project context, then either let Aito draft the full plan in one shot — or build it step by step, asking Aito for the next prediction at each click.
Operations›Project Plan
_predict + _recommend
tasks
Tables
assignee · days · cost · success
Targets
history
Sources
_predict_recommend
POST /api/v1/_predict
{
"from": "tasks",
"where": {
"project_type": "construction",
"phase": "mep",
"task_name": "HVAC commissioning",
"region": "Helsinki",
"season": "summer"
},
"predict": "subcontractor"
}
{
"from": "tasks",
"where": {
"project_type": "construction",
"phase": "mep",
"task_name": "HVAC commissioning",
"region": "Helsinki",
"season": "summer"
},
"predict": "subcontractor"
}
How it works
Project Plan combines three Aito flows. The generative side runs aito.._predict three times per task (assignee_kind, subcontractor or assignee_person, planned_days, planned_cost_eur) given the task context — drafting a complete plan from precedent instead of a hand-coded template. For each phase it also runs aito.._predict on the purchases table to auto-draft material POs to the supplier the buyer's history points at (the Lemonsoft+Jakamo punchline — Aito routes the spend before anyone touches the requisition). Click any task row to swap to the matchmaking flow: aito.._recommend with goal: {success: true} ranks subcontractors by predicted P(success) for that exact (phase, region, season) context.
Learn more