What does it actually take to build a tech platform for preconstruction? Gareth McGlynn puts that question to Peter Alderson, Co-Founder and CEO of Province, who spent a decade building technology with his co-founder Ian before turning to construction, where Province is now doing very well in the multifamily space.

Discussion Highlights:

• All About Province: The company Peter co-founded with Ian, and how it's doing very well in the multifamily space.
• Peter's Path: His inspiration for going into construction, then picking preconstruction.
• Building a Tech Platform in Preconstruction: The pitfalls and what it's actually like to build one.
• AI in Preconstruction: The wonderful world of AI and where it actually fits.
• New Product Release: Province has a new product and Peter shares it right on the episode.

You can connect with Peter via his Linkedin: https://www.linkedin.com/in/peteralderson/

You can also find Province at https://www.getprovince.com/ or reach Peter directly at [email protected].

Deepti Yenireddy, Founder and CEO of Boon, joins Gareth McGlynn to talk about building AI that can actually read construction drawings. A second-time founder, Deepti exited her first company, HR tech platform My Ally, and led product at IoT leader Samsara before turning her focus to preconstruction.

Key Topics Covered:

• What Boon Does: Helping preconstruction teams bid more, bid better, and bid more accurately, at 10x to 100x speed.
• The AI Estimator: Deepti announces Boon's first preconstruction AI employee, working the full lifecycle from project invite to final proposal.
• Hallucination and Human Control: How estimators stay in the loop with a visual review platform and links back to the agent's sources.
• Accuracy and F1 Scores: Why Boon publishes accuracy scores for every trade and scope, and why below 90 percent creates user friction.
• Boon vs Generic LLMs: Why ChatGPT and Gemini can read the text in your drawings but cannot tell where a line starts and stops.
• Implementation and ROI: A one-hour onboarding, and a GC customer who generated 430 extra hours in a year with one person using the agent.
• And much, much more.

You can connect with Deepti via her Linkedin: https://www.linkedin.com/in/deepti-yenireddy/

Wharton Smith Senior Estimator Zackery Batt joined Gareth McGlynn at Advancing Preconstruction 2026 in Phoenix. A carpenter by trade, Zackery went from using no AI at all to running a custom preconstruction AI system built inside Claude Cowork, all in the space of nine months.

Discussion Highlights:

• From Carpenter to AI Builder: How spotting a LinkedIn post from Turner Construction's head of AI development pushed him to teach himself AI through YouTube and six months of trial and error.
• Bid Leveling, Almost Automatic: Saving every subcontractor's exclusions and assumptions so the system flags anything outside their normal boilerplate.
• Scope Generation at 95% Accuracy: The trickier workflow to crack, from early chats that hallucinated scope to a system that now finds almost everything he looks for.
• Knowledge Retention in MD Files: Years of carpentry, flooring, and millwork knowledge written into files Claude reads before generating scope, picking up items that are technically there but not actually on the drawings.
• Client Proposals in Two Minutes: How AI pulls awarded subcontractor proposals and generated scope into a 10 to 15 page client-facing proposal that used to take five hours.
• Where It's Heading: One estimator managing multiple AI workflows instead of five or six, the adoption hurdles that will hold AI back in construction, and why the companies that own their data will thrive.
• And much, much more.

You can connect with Zackery via his Linkedin: https://www.linkedin.com/in/zackery-batt-98643a393/

Spencer Thompson, Lead Estimator at Cardinal Civil Contracting, joined Gareth McGlynn at Advancing Preconstruction 2026 in Phoenix, Arizona. As a self-performing heavy civil contractor operating across the Carolinas, Cardinal Civil works in some of the most geologically unpredictable terrain in the country.

Key Topics Covered:

• Earthwork Pitfalls: The lack of geotechnical report data as the biggest risk in civil estimating, and how soil types can change completely from one site to the next on the same road.
• Early Involvement: Why bringing the civil contractor in at the early design phase allows the team to fit the design around the site, not the other way around.
• Retaining Walls: How placement near property lines and the need to import specific backfill material can create major unplanned cost and scheduling impacts.
• AI in the Workflow: Using Microsoft Copilot with EIA diesel price data to build fuel escalation models and present projected cost scenarios to clients.
• And much more.

You can connect with Spencer via his LinkedIn: https://www.linkedin.com/in/spencer-thompson-a0924690/

More conversations like this one are on the way. Subscribe to The Preconstruction Podcast so you don't miss what's next.

Marc Apicella, Co-founder and CEO of Belidor, joins Gareth McGlynn at Advancing Preconstruction 2026 in Phoenix. A product designer by background with graduate research in recommendation systems at Google DeepMind, Marc is building AI-powered tools to take the friction out of scope generation and bid leveling in preconstruction.

Discussion Highlights:

• Founding Story: A GC brother in upstate New York exposing the real pains of how the industry works, and why that became the foundation for Belidor.
• What Belidor Does: Scope generation from specs and plans in under 10 minutes, Building Connected integration, and bid leveling across any format.
• Knowledge Retention: Why the retirement wave is the core problem Belidor is solving, and the 30-year estimator hired back at 30% more as the reality check.
• Three Product Promises: User friendly, outputs linked to source documents, and reliable on bid day.
• Handling Change Resistance: Why estimating takes longer to master than almost any profession, and how Belidor builds around existing workflows rather than disrupting them.
• Data & the Five-Year Vision: SOC 2 compliant, historical data shareable to supercharge the platform, and a longer-term goal of going from prompt to ready-to-build in days to weeks.

You can connect with Marc via his LinkedIn:  https://www.linkedin.com/in/marcapicella/
Or reach him directly at [email protected]

Stay connected with The Preconstruction Podcast for more conversations with the innovators shaping the future of preconstruction. Don't forget to subscribe!

Gareth McGlynn sits down with Luigi La Corte, Co-Founder and CEO of Provision, recorded live at Advancing Preconstruction 2026 in Phoenix. Provision is an AI platform built to solve scope gaps in preconstruction, a problem La Corte describes as costing GCs an estimated 0.5–2% of total project cost.

Key Topics Covered:

• Career Path to Provision: From civil engineering at Arup, to infrastructure finance at Plenary, to founding Provision after observing scope failures at the largest GCs in North America.
• Provision's Pivot: From contract review to Scope Agent, targeting scope gaps after general AI models made document review defensible.
• Why Scopes Are Hard for AI: LLM limitations with image interpretation, thousands of edge cases around drawing legends and plan references, requiring purpose-built tooling.
• $1M Dataset Investment: Hundreds of thousands of estimator-verified line items across real projects, built to substantiate accuracy claims.
• Real-World Results: One customer from three weeks to four days on scope prep. Addenda changes caught before subcontractor distribution, preventing a six-figure issue.
• AI and the Estimator: Automating rote workflows (spec review, RFI conflicts, scope extraction) without replacing the estimator.

You can connect with Luigi via his LinkedIn: https://www.linkedin.com/in/luigi-r-la-corte/

Wyre AI Founder and CEO Sunil Dorairajan sat down with our host, Gareth McGlynn, at Advancing Preconstruction 2026 in Phoenix, Arizona. A construction project manager turned entrepreneur, Sunil brings a contractor's perspective to AI-powered preconstruction tools.

Key Topics Covered:

• From Field to Founder: His ten years in heavy commercial construction, the founding of Pipe in 2017, its acquisition by Autodesk, and what brought him back to the startup world with Wyre AI.
• Why Preconstruction, Why Now: The combination of team burnout and a generation of retiring estimators creating urgency for technology adoption in precon.
• WireScopes: How it takes a week-long scope production process and brings it down to a matter of hours, and what that means for precon teams under tight bid deadlines.
• WireCheck: How it catches conflicts between plans and specs before a bid goes out, and a real contractor example where a single discrepancy flagged pre-bid was worth $25,000.
• AI Without the Complexity: Why Wyre AI is model-agnostic, what that means for estimators, and why precon managers shouldn't need to be model mechanics.
• And much, much more.

You can connect with Sunil via his LinkedIn: https://www.linkedin.com/in/sunildorairajan/

You can also reach Sunil directly at [email protected] or visit www.wyreai.io to request a demo.

Gareth McGlynn speaks with Nate Weeks, Director of Estimating at Jewett Construction, a commercial general contractor based in Southern New Hampshire with operations across New England, upstate New York, and into the mid-Atlantic. Jewett has been in operation for over 50 years and remains family-owned into its second generation.

Discussion Highlights:

• Career Progression: From zero estimating background at 27 to Director of Estimating in under a decade.
• The Shift to Leadership: Setting clear expectations, delegating effectively, and developing the team.
• Planning as a Sales Tool: Using a dedicated planning manager and Giraffe software to help developers optimise land use before estimating begins.
• Tech Stack: Builder, HubSpot, Building Connected, Procore, and Excel for estimating.
• The Bid Leveling Gap: Data entry from subcontractor bids as the biggest time drain, and what a real solution needs to look like.
• AI in Preconstruction: Using AI as a research aid while staying cautious about its reliability for cost decisions.

You can connect with Nate via his LinkedIn: https://www.linkedin.com/in/nate-weeks-2409385a/

More great conversations ahead. Stay connected with The Preconstruction Podcast for weekly episodes featuring voices from across the industry.

Gareth McGlynn speaks with Nathan Schafer, Estimating Manager at Cornerstone General Contractors, at Advancing Preconstruction 2026 in Phoenix. A self-performing GC based in Alaska, Cornerstone gives Nathan a hands-on perspective on preconstruction that is grounded in real field conditions.

Key Topics Covered:

• Nathan & Cornerstone: Working across healthcare, hospitality, and federal military projects, with a focus on vertical commercial construction.
• New Estimators & Value Engineering: The shortage of talent entering the field, their understanding of value engineering, and professional development through ASPE and AACE.
• Lean Construction: Its growing impact on the estimating process and the role of AI takeoff tools as a lean principle in action.
• Data Science in Preconstruction: How Cornerstone is incorporating data science into its workflow, including labor productivity tracking and predicting quantity growth risk as a function of design maturity.
• Value Management in Practice: Cornerstone's formal pilot on a $90M project: 13 propositions totaling $9M, with $7M accepted.

You can connect with Nathan via his LinkedIn: https://www.linkedin.com/in/nathan-schafer-cpe-b75991178/

Or reach him through his blog: https://www.preconomics.com/blog

Recorded live at Advancing Preconstruction 2026 in Phoenix, Arizona, this episode features Steve Dell'Orto, Founder and CEO of Concentric.

A construction industry veteran who built the platform from the GC side, Dell'Orto brings the perspective of someone who has lived the inefficiencies he is now working to solve.

Key Topics Covered:

• The Excel problem: Most precon teams still running disconnected spreadsheets with no standardization, structured data, or shared repository across the team
• Platform adoption: Concentric is designed to be familiar and Excel-like, keeping the barrier low without reinventing existing preconstruction workflows
• The data flywheel: Each project captured in the platform feeds the next, building accuracy, precision, and certainty of outcome over time
• Single pane of glass: Consolidating precon workflows into one platform eliminates the data transfer friction and security exposure that comes with juggling multiple point solutions
• Amplify and agentic AI: Concentric's AI layer operates inside the platform vault, keeping proprietary data secure while delivering company-specific, context-rich outputs
• Precon as a unifier: The platform is designed to bring operations into the planning stage earlier, bridging preconstruction and execution rather than treating it as a clean handoff

You can connect with Steve via LinkedIn and book a demo directly through the Concentric website:

• LinkedIn: https://www.linkedin.com/in/stevedellorto/
• Website: https://concntric.com/