ABEPIM — The AI-Native OS for Commercial Roofing
Commercial roofing contractors stitch together 5–8 disconnected tools to run their business. Estimation alone takes 4–12 hours per project. The $92B U.S. roofing market (all segments; commercial share requires separate market sizing) has 100,000+ contractors — and no end-to-end AI-native platform serving commercial workflows. Point tools exist (On Center, BuildVision AI); no full AI-native commercial OS is known to us. Residential is solved (Hover, Roofr, EagleView). Commercial is wide open.
ABEPIM is a vertical operating system for commercial roofing contractors — AI-powered estimation, proposal generation, project management, job costing, and autonomous back-office operations. Working prototype operational on live projects today. Built on AI models — production roadmap migrates from today's brokered APIs to self-hosted open-source models on owned hardware, targeting an order-of-magnitude compute-cost advantage over frontier-API competitors and keeping client blueprints on infrastructure we control. The AI engine is trade-agnostic — roofing is the beachhead, with expansion to HVAC, electrical, and plumbing opening a multi-trade expansion path across all MEP trades.
| Company | Focus | Valuation | Commercial Roofing? |
|---|---|---|---|
| Procore | GC construction management | $9.2B (public) | No — built for GCs |
| ServiceTitan | Home services OS | $8.9B (IPO) | No — residential |
| BuildOps | Commercial MEP | $1.0B (Series C) | No — HVAC/plumbing |
| AccuLynx | Residential roofing | $2.35B (acq. offer) | No — residential only |
| ABEPIM | Commercial roofing OS | Seed stage | First end-to-end AI-native |
- Working prototype on live commercial projects — e.g., 20 sheets deep-extracted from a real 120-pp bid set in ~28 min of AI processing (expert review follows)
- Pilot customer generating revenue (15% platform fee)
- Tampa Bay Wave CORE — July application history; no current acceptance claimed with updates
- SCORE Mentor Program accepted
- New World Angels application submitted ($250K / $2.5M cap)
- Senior Advisor: Dan Cain (ex-SAS VP/GM, Reef CSO)
- Jason van Joosten — Founder/CEO. USAF veteran (SBA VetCert pending). 10+ yrs roofing industry experience.
- Wesley Schlemmer — Technical Lead. AI/ML architect. Built ABEPIM prototype. B.S. Tufts. Cory Williams — Owner, America's Best Roofing and Construction; Operations Partner in the ABH structure. Pilot customer. Field validation. (He is not the licence holder — Talli Houston holds CCC1332167.)
- Dan Cain — Senior Advisor. CSO Reef CyberSecurity. Ex-VP/GM SAS (6 yrs).
- Nashville Automation Co. — Technical Build Partner. AI engineering firm, 8+ clients. Construction AI deployments. LOI signed.
"As a Florida-licensed roofing contractor with 10 years in the field, I've seen every tool promise to 'revolutionize' estimating. ABEPIM is the first one that actually delivers. It pulls measurements, calculates materials, and generates a complete proposal in minutes — what used to take my team days. ABEPIM isn't just software — it's the difference between us bidding on a handful of jobs a month and bidding on twenty-plus. I've been doing this long enough to know when something is a gimmick and when something is a game-changer. This is a game-changer."
— Cory Williams, Owner, America's Best Roofing and Construction (FL License CCC1332167)
Six compounding barriers, each 12–24 months to replicate: (1) Harness engineering — the six-layer execution system around the model (task contract, context compiler, permissioned tool gateway, durable state, evidence gates, trace and recovery) that turns model output into verified takeoffs; a competitor can copy a prompt, not six layers of enforcement. (2) Training data flywheel — every project makes the platform smarter; competitors start at zero. (3) Cost architecture (roadmap) — self-hosted AI targets a step-change lower compute cost than cloud-API competitors. (4) Code compliance engine — Florida HVHZ regulatory moat. (5) Domain expertise — built by operators who run a roofing company daily. (6) Brand + customer lock-in — first-mover with an end-to-end AI-native platform, trademark filed.