AI for project controls
Practical AI workflows I build to take the grind out of planning and controls — turning raw site inputs, cost workbooks and schedules into checked, report-ready outputs in minutes, not days. A human always reviews before anything is issued.
DPR Generator
Converts raw site inputs — progress notes, manpower counts, equipment logs — into a clean, formatted Daily Progress Report ready for client submittal.
Site notes, manpower, equipment, weather, photos
Structures, summarises & flags missing fields
Formatted DPR + carry-forward to WPR
WPR vs DPR Checker
Verifies that the activities planned in the weekly programme actually match what the daily reports show on the ground — and surfaces the gaps.
Weekly plan (WPR) + the week's daily reports (DPR)
Reconciles planned vs actual activity by activity
Variance list + slippage & catch-up flags
Contract Scope Verifier
Checks additional or instructed site works against the contract scope, so potential variations and claims are spotted early instead of absorbed silently.
Contract scope / BOQ + site instruction or extra work
Matches against scope; classifies in/out of contract
Variation candidates + clause references to review
Budget Workbook Cleaner
Cleans and restructures messy cost-control workbooks — consolidating cost codes, fixing inconsistencies and shaping the data so dashboards can read it.
Raw / inconsistent cost workbook
Normalises codes, dedupes, validates totals
Clean, dashboard-ready cost dataset
Schedule Audit Assistant
Reviews a P6 schedule for logic, links and resource assumptions — flagging open ends, negative float, constraints and broken logic against DCMA-style checks.
P6 export (XER / XLSX) of the schedule
Runs logic / float / constraint health checks
Audit report with prioritised fixes
A note on responsible use: these workflows assist a qualified planning engineer — they do not replace one. Every output is reviewed by a human before it is issued, and no confidential client data is shared in the samples shown across this site.