[ Case Study · Construction ]

Kilcor Painting & Coatings runs change-order phase coding 55% faster

Kilcor Painting & Coatings, an employee-owned commercial painting and coatings contractor, now runs change-order phase coding 55% faster and starts preconstruction with an indexed drawing set instead of reading the spec book by hand. Fastlane AI mapped where AI could help and rolled Claude out company-wide. It shipped 11 AI capabilities, Kilcor staff built seven more, and 91% of the organization adopted AI in their daily work within five weeks.

change-order phase coding
55% faster
Client-reported
AI use cases live across 4 functions
18
Client-reported
AI adoption across the organization in five weeks
91%
Client-reported

01The problem

Kilcor Painting & Coatings is an ENR Top 20, employee-owned commercial painting and coatings contractor headquartered in Alpharetta, Georgia, with offices across Georgia, Alabama and Texas. Its work includes hyperscale data centers and manufacturing facilities.

Much of the work ran by hand. Estimators read hundreds of pages of drawings and specs before takeoff. Project managers built a phase-code breakdown in Excel for every change order, 15 to 20 minutes each on top of pricing. Offer letters were rebuilt from the previous hire’s document.

AI use was sporadic, with no shared workspace or standard.

02The work

Fastlane AI started with an AI Discovery Sprint. Its Discovery Agent™ interviewed team members across seven functions, from superintendents to the President and the CFO, and Fastlane AI turned the transcripts into workflow maps of where AI could change each critical-path process.

EDGE came next. Fastlane AI deployed Claude to the whole company, configured Connectors, and built Skills, Projects and Cowork Agents against the highest-value opportunities in estimating, project management, field operations, FP&A and HR. It shipped 11 purpose-built AI capabilities against the discovery findings. In finance, AP and AR aging cleanup, COI compliance checks and job setup checks run against real Spectrum exports and executed contracts. In HR, an offer letter capability computes every derived figure from the intake form.

03How people took it up

Training opened with company-wide Foundations sessions, then role-specific cohort training for each department, so people could use the tools configured for them and build their own.

They built. Kilcor staff made seven AI capabilities of their own and shared them company-wide, and a project manager built the Claude Skill that now handles change-order phase coding. Within five weeks, 91% of the organization had adopted AI in their daily work, from superintendents through the CFO.

04The outcome

Preconstruction used to start with an estimator reading a 400-page spec book by hand. It now starts with an indexed drawing set and the cost-driving notes already surfaced, so a fixed estimating team can pursue more bids without adding headcount. Eighteen AI tools are live across four functional areas.

Change-order phase coding runs 55% faster, 7 to 8 minutes instead of 15 to 20, validated on real jobs, which moves approved work to billing sooner. WIP-to-CRM reconciliation went from 2 hours per run to 15 to 20 minutes.

What we built

Project management

  • Before

    Project managers split each change order by phase code in a spreadsheet, by hand.

    After

    A project manager drops in the change-order PDF and gets the phase-code sheet back.

Estimating

Field operations

Finance and accounting

Compliance and risk

Operations

Fastlane showed us where the manual hours in our business actually go. They gave us a ranked map of the manual work across the business, and deployed AI against critical path processes.
Adam DiamondCEO, Kilcor Painting & Coatings

Questions about this work

How long did it take?

The work began with an AI Discovery Sprint and continued through EDGE. Adoption reached 91% of the organization in five weeks.

Who was involved?

The Discovery Agent™ interviewed team members across seven functions, from superintendents to the President and the CFO. Training reached every department, first in company-wide sessions and then in role-specific cohorts.

Did the team build anything themselves?

Yes. Kilcor staff built seven AI capabilities of their own and shared them company-wide. A project manager built the Claude Skill that produces change-order phase-code breakdowns.

Applied AI custom built for your organization.

Fill out the form to connect with our team, and an AI strategist will be in touch within 24 hours.

Required

Kilcor case study: 55% faster phase coding | Fastlane AI