SNC Discovery Session — Summary & Opportunities
Prepared by James Andrew Walsh, Amplify Luxury
February 19, 2026
Discovery
Meeting Overview
On February 18, 2026, James Walsh of Amplify Luxury met with Marc Markwell (CFO), Bryan (VP), Moe (IT Manager), and Randy McReynolds of Artistry Building Group at SNC's office in Sparks, Nevada. The purpose was to explore how AI-powered automation could address operational challenges across SNC's business.
The conversation was wide-ranging and substantive, covering estimating, field operations, payroll processing, equipment management, and sales. What emerged was a clear picture of a well-run company with deep institutional knowledge — and significant opportunities to turn that knowledge into automated, real-time operational intelligence.
What We Heard
Equipment Utilization: Millions Left on the Table
SNC operates a mixed fleet of owned and rented equipment, with roughly half the fleet being rentals at $2,000–$3,000 per day. Equipment telemetry exists through E360 and manufacturer systems, reporting utilization, idle time, and hours running.
The problem: this data is only reviewed monthly — and even then, it's six weeks after the end of the previous month. By the time anyone sees that a piece of equipment sat idle for a week, the money is already spent and the conversation is stale.
The gap between what timecards report and what the equipment actually ran is significant. In one example discussed, rental invoices showed 20 hours of runtime for a month while timecards reported 160 hours. Multiple foremen sometimes log the same piece of equipment on their timecards, inflating reported hours from 40 to 160 in a week.
Bryan put it plainly: "There's so much money here. We're talking millions."

The opportunity: Real-time comparison of E360 telemetry against timecard-reported hours — daily, not monthly. Same-day alerts when equipment is sitting idle, being double-reported, or underutilized. Decisions made with current data instead of data that's six weeks old.
The Payroll Audit Pipeline: Manual Verification at Scale
Every week, timecard data flows from Dispatch to HeavyJob to Spectrum through a multi-step audit process involving four to five people. They're checking for:
  • Self-consistency (do start and end times add up to reported hours?)
  • Correct pay scales (blade operator vs. excavator operator)
  • No double-counting of employees across timecards
  • No one showing 16-hour days from being on two timecards
  • Correct cost code assignment
  • Equipment not appearing on multiple foremen's timecards simultaneously
Bryan, the fastest person in the company at this review, spends three hours on it. Payroll editing alone consumes over 10 hours per week. The full audit chain involves superintendents, foremen, project managers, and two payroll staff — all looking at the same data for different types of errors.

The opportunity: Automated validation rules that catch these discrepancies the moment a timecard is submitted — before it reaches payroll, before it reaches accounting. The same checks these five people perform manually, running continuously and flagging only what needs human attention.
Field Sales: Three Days When It Should Be Thirty Minutes
Bryan described a field sales program he built 30 years ago that allowed sales reps to generate proposals on-site in 30 minutes. The concept was sound: input the square footage, number of islands, distance from the hot plant, product type, and area-based pricing — and hand the customer a proposal before leaving.
That program was never adopted at SNC. Today, the process takes three days: a field rep writes up the details, hands it to someone who prices it, who hands it back, and then the rep goes to sell.
Bryan's cardinal rule: "Never leave the customer. Finish what you're doing, sell it, hand them a proposal, get the sale."

The opportunity: An AI-powered field sales tool where a rep describes the job verbally — "45,000 square feet of type 2 slurry, five parking islands, located here" — and receives an email-ready proposal within minutes, priced from SNC's current rate structure. The sale happens on-site, same day.
Plan & Specification Review: Hours of Searching
When SNC receives plans and specifications for a new job, someone sits down and reads through them page by page, looking for specific items: railroad insurance requirements, special conditions, key details, and potential omissions.
Marc described a common scenario: a note on the plans references a masonry wall, but there's no corresponding detail. It gets missed. Then out on the job, the crew discovers the requirement — and it's too late to account for it in the bid.
The process is done both on paper (Darcy prints and reviews page by page) and on screen (Sarah and the estimators review PDFs). Either way, it's hours of manual searching for information that should be instantly identifiable.

The opportunity: Automated plan scanning that extracts key requirements, flags missing details, identifies notes without corresponding specifications, and creates a structured checklist of everything the estimating and operations teams need to know — in minutes instead of hours.
Estimating: Two Worlds, One Goal
SNC's estimating work falls into two categories:
Public works
Plans and specs are provided with defined quantities. This is more structured and repeatable. Quantities should flow directly into the bidding software, with crews auto-assembled based on historical patterns. Four full-time estimators handle this work using SharpeSoft Estimator and AGTEK for earthwork takeoffs.
Private works
Plans are sometimes only 30% complete when SNC is asked for a budget. The team builds quantities from incomplete information, and the scope shifts as plans develop. Marc described it as "building the plane as we take off from the ground."
On the public works side, Bryan identified the cookie-cutter opportunity: jobs that repeat with variations in depth, speed, and scale. These follow predictable patterns and shouldn't require an estimator to start from scratch each time.
Martha operates AGTEK for earthwork takeoffs — a system James has deep personal experience with, having spent thousands of hours on AGTEK digitizer boards doing earthwork takeoffs as a teenager for his father's construction estimating company.

The opportunity: For public works, automated extraction of bid item quantities from plan specifications directly into SharpeSoft's import format — reducing the manual entry that consumes estimator time. For private works, AI-assisted quantity building from partial plans, with clear flags on assumptions and missing information. A meeting with the estimating team (Mike and Martha) would define the specific workflows where automation has the highest impact.
Disconnected Systems: Data Exists, Decisions Don't
A theme ran through every topic: the data SNC needs to make better decisions already exists — in HeavyJob, in Spectrum, in E360, in Dispatch, in SharpeSoft. The challenge is that these systems don't talk to each other in real time.
Bryan summed it up: "You can't make decisions because you don't have the data. Even though the data exists in these different places, you can't put it together."
Equipment runs for a week while no one knows it's idle. Timecards go through five layers of manual review because there's no automated cross-check. Sales proposals take three days because the pricing lives in one place and the field data lives in another. Production quantities go untracked because the team is too short-staffed to measure what they should be producing.

The opportunity: A connected intelligence layer that bridges these systems — pulling data from where it lives, applying the business rules SNC's experienced people already know, and surfacing the right information to the right person at the right time. Not replacing the expertise, but giving it real-time data to work with.
Themes
Four patterns emerged from our conversation:
1
Data exists but isn't connected
Every system has valuable information. The gap is in bringing it together for real-time decision-making.
2
Decisions lag weeks behind reality
Monthly reviews and six-week reporting cycles mean problems are discovered long after the money is spent.
3
Senior people are consumed by manual verification
Four to five people auditing timecards, estimators entering quantities by hand, field staff flipping through plan pages — experienced professionals doing work that should be automated.
4
Operational excellence is the competitive advantage
SNC has strong leadership, deep experience, and a culture of doing things right. The opportunity isn't to change how SNC works — it's to give SNC's people the tools to work at the speed their expertise deserves.
Suggested Next Steps
01
Meet with the estimating team
A conversation with Mike and Martha to understand the specific estimating workflows, AGTEK usage, and SharpeSoft processes where automation would have the most impact.
02
James to prepare a detailed proposal
Focused on the highest-impact opportunity identified in this session, with a clear scope, timeline, and deliverables.
We're excited about what's possible here. The combination of SNC's operational depth and AI-powered automation has the potential to set a new standard for what a construction company can achieve.
— James Walsh, Amplify Luxury