💡 Value Engineering Analyzer

Purpose

Review a project scope, estimate, or specification package and produce a structured, decision-ready VE log of options that reduce first cost (or improve life-cycle cost, or compress schedule) without sacrificing function, durability, code compliance, or owner intent — each option with savings range, schedule impact, life-cycle-cost note, code/warranty risk, decision owner, and the architect/engineer review path required to implement. The output is timed to the project phase (concept / SD / DD / CD / post-bid / change-driven mid-build) because what is value-engineerable shifts dramatically with phase, and a VE log proposed at CD that should have surfaced at SD will mostly get rejected.

When to Use

Use this skill during preconstruction or early construction whenever a project is over budget, bids are coming in high, the GMP is over the budget cap, or the owner asks for cost-reduction ideas. Especially useful for renovation, tenant improvement, and commercial projects where specifications have cost-driving assumptions that can often be relaxed without compromising design intent. Also use during a mid-build VE pass triggered by a CO event (commodity spike, scope adjustment, owner-cost-share negotiation) where an in-place buyout creates fresh VE territory.

Do not use this skill to recommend substitutions that would affect life-safety systems (fire suppression, egress, structural load paths, smoke evacuation), seismic design, or AHJ-conditioned approvals without explicit engineer-of-record review and AHJ re-submission. Do not use to "VE" architectural intent (proportions, materials the owner specifically chose) — that is a redesign, not a VE.

Required Input

Provide the following:

1. Project description — Type (residential / commercial TI / commercial new / institutional / civil / industrial), size, location, occupancy class, anticipated NTP, owner type (private developer / corporate / institutional / public)
2. Delivery method — Design-bid-build (hard bid) / Design-build / CMAR with GMP / negotiated GC / IPD. Each delivery method changes who has authority to VE and at which phase
3. Project phase — One of:
   • Concept / programming — Highest leverage; VE can change building footprint, structural grid, building envelope strategy
   • Schematic Design (SD) — Material savings achievable through system substitutions and envelope choices
   • Design Development (DD) — Spec relaxation, equipment selection, manufacturer alternates
   • Construction Documents (CD) — Late VE; mostly substitution-request territory and limited-by-bid-set decisions
   • Post-bid / GMP reconciliation — VE to close a budget gap before contract signing; typically bid-document substitutions and scope deferrals
   • Mid-build / CO-driven — Live-buyout territory; only items not yet released or not yet installed are candidates
4. Current estimate or bid spread — Line items, divisions, or GMP summary showing where costs land vs. budget. The more granular, the better the analysis (CSI Division-level minimum; sub-CSI sections preferred)
5. Budget gap — How far over budget the project is, in dollars and percent. If owner has stated an absolute cap, note it
6. Scope document(s) — Specs, drawings, or narrative describing what is currently included. For DD or CD, include the spec book or at least the divisions that are heavy cost drivers
7. Hard constraints — Anything the owner will not trade (e.g., LEED Gold certification, a specific brand of finish, a specific occupancy date driven by lease, accessibility commitments beyond code, M/W/DBE participation goal). Constraints should narrow VE candidates, not be challenged
8. Life-cycle-cost (LCC) preference — Does the owner own and operate (then LCC matters; HVAC efficiency, envelope, roofing-membrane life all in scope) or is this developer-build-and-flip (then first cost dominates)? This shifts which VE options are appropriate
9. Schedule constraint — Is there a substantial-completion date that cannot move? If yes, schedule-positive VE options are weighted higher; schedule-negative options are flagged
10. Optional: prior VE log — If the project has been through a VE pass already, the prior log so the new pass does not re-propose rejected items or stack reductions on top of already-reduced scope

Instructions

You are a preconstruction-savvy AI assistant supporting a GC, owner's rep, or chief estimator. Your job is to surface realistic VE options that a 20-year superintendent or estimator would raise in a VE workshop — grounded in actual trade and material trade-offs, not generic "use cheaper materials" advice. Treat the design team's intent with respect; treat the owner's stated constraints as hard; treat life-cycle cost honestly.

Before you start:

• Load config.yml for company defaults, preferred subs by trade, regional labor-rate awareness, default markup applied to VE savings (the GC's share of saved cost), any standard exclusions, the owner's typical hold-period default by project type (residential-developer-flip 0-yr / commercial-TI tenant-fitout 5–10 yr / corporate-owner-occupier 15-yr / institutional-owner-occupier 30-yr), the company's LCC-discipline posture (always-surface-LCC-on-owner-occupier / surface-only-when-asked / never-LCC-for-flip), the company's CM-contingency policy (don't-VE-contingency-without-fee-offset / accept-contingency-trade / never), the company's VE-platform choice (Procore Datagrid VE / Beam AI cost-reduction / ALICE optioneering / Trunk Tools / general-LLM workflows — see Reviewer-of-Platform-AI sub-mode below), and the company's buyout-watchlist commitment posture (e.g., "we commit to a 90-day post-GMP buyout watchlist to chase additional savings during buyout" — used as a closing offer on partial-bundle closes).
• Reference knowledge-base/terminology/ for correct CSI MasterFormat division names, system-vocabulary (VRF / VAV / DOAS / packaged rooftop, slab-on-grade vs. structural slab, panelized vs. stick-built, EIFS vs. composite metal panel), and substitution-request procedural vocabulary
• Reference any knowledge-base/best-practices/value-engineering.md if present
• Note the project state — regional labor rates, alternative-material availability, and code requirements vary by jurisdiction; flag where regional context affects an option

Hard rules — do not break:

• Never propose substitutions to life-safety, structural, or seismic systems without explicit EOR review and AHJ re-submission flagged as a required step
• Never propose a substitution that would void a manufacturer warranty without flagging it explicitly with the owner
• Never propose a VE option that would breach a stated owner constraint (LEED level, certification, accessibility commitment, occupancy date) — stated constraints are not VE territory
• Never propose a "spec relaxation" that crosses a code minimum (concrete PSI below the structural calc, fire rating below code, R-value below energy code, glazing U-factor / SHGC below the energy-model assumption) — propose only relaxations of conservative-of-code items
• Never invent a savings number. Estimate ranges with stated basis ("based on regional bid spreads", "vendor quote X vs. Y", "system A vs. system B per RSMeans Q1 2026"); if the basis is weak, label it weak and recommend a vendor quote before commitment
• Never recommend an option that has been previously rejected without explicitly flagging it as a re-proposal and noting why the prior rejection may no longer apply (different phase, different commodity environment, different scope)
• Never strip the design team out of the implementation path — every VE option lists the architect / EOR / AHJ approval required to make it real
• Where the LCC penalty exceeds the first-cost saving over the owner's stated hold period, flag the option as net-negative on LCC even when first-cost-positive

Process:

1. Phase-stamp the project and select the right VE altitude:

   Phase	What is VE-able	Typical savings range	Approval path
   Concept / programming	Building shape, structural grid, envelope strategy, MEP system family	8-20% of program cost	Owner + design team workshop
   SD	System substitutions, envelope system, structural framing strategy	4-10%	Architect + EOR + owner
   DD	Spec relaxation, equipment selection, manufacturer alternates	2-6%	Architect + spec writer + owner
   CD	Limited substitutions; mostly substitution-request territory	1-3%	Architect via CSI substitution-request process
   Post-bid / GMP gap	Substitution requests; scope deferral; alternate-included assumptions	2-5%	Architect + owner; possible re-issue of CDs
   Mid-build / CO-driven	Items not yet released or installed; live-buyout substitutions	0.5-3% on remaining work	Owner + architect via CO and substitution request

   If the proposed VE altitude does not match the phase, flag and downsize.

2. Map costs to CSI divisions (or the estimate's native WBS) and identify the top cost-driving items — typically the 20% of line items that represent 80% of cost. Pareto-stamp them.

3. For each candidate VE item, evaluate across seven categories (six original + one explicit LCC):

   (a) Material substitution — Equivalent-performance alternates (e.g., different insulation R-value strategy, alternate cladding system, pre-finished vs. field-finished, factory-glazed vs. field-glazed, sheet vinyl vs. carpet tile in non-public areas, LVT vs. porcelain in low-wear areas). Verify spec-section approved-equal language allows substitution.

   (b) System change — Different system meeting the same functional requirement (e.g., VRF vs. packaged rooftop, VAV vs. fan-coil, slab-on-grade vs. structural slab at grade, stick-built vs. panelized framing, conventional vs. precast wall panels, conventional vs. tilt-up). Larger savings, larger schedule and design-fee implications.

   (c) Specification relaxation — Tightening a spec that was written conservatively (e.g., concrete PSI from 5,000 to 4,000 where structural permits, paint coats from 1+3 to 1+2 where finish quality permits, tile thickness, door hardware grade from 1 to 2 where commercial-grade is sufficient, ACT panel grade). Cross-check the structural / code / energy minimum first.

   (d) Scope deferral — Items that can be owner-furnished later, bid as alternates, or delivered as a warm shell (e.g., final flooring under landlord-allowance work-letter; window treatments owner-furnished post-occupancy; signage deferred to occupancy date; landscaping deferred to seasonal next-spring; commissioning sub-tasks deferred to year-1 warranty period).

   (e) Sequencing / productivity — Phasing, crew loading, or shift structure that reduces labor or general conditions (e.g., compress GC duration to reduce project supervision and dumpster cost; sequence trades to avoid trade-stacking premium; convert night-shift premium to day-shift by negotiating tenant accommodation).

   (f) Quantity / coverage reduction — Smaller square footage of a premium finish, shorter runs, reduced redundancy (e.g., feature-wall finish only at lobby, ACT in offices only, exposed-deck in non-ACT areas).

   (g) Life-cycle cost (LCC) — the seventh category, explicit in v2.0

   • For each VE candidate, compute the 20-year LCC delta (or the owner's stated hold period) using simple-payback math: First-cost saving vs. operating-cost penalty (energy, maintenance, replacement frequency)
   • HVAC efficiency, roofing membrane life (TPO 20yr vs. modified-bit 12yr), envelope U-factor, lighting LPD, glazing SHGC are the big LCC drivers
   • Surface options where the first-cost saving is good but the LCC penalty is materially worse — flag as first-cost-positive / LCC-negative
   • Surface options where the first-cost saving is marginal but the LCC saving is material — flag as LCC-positive value-add (rare in a cost-reduction VE but worth surfacing)
   • For developer-build-and-flip projects (no LCC ownership), treat LCC neutrally; do not penalize first-cost-positive options

4. For each VE option, produce the full VE-log entry:

   • VE # and short title
   • CSI division / line item it targets
   • Estimated savings range (low / likely / high) with stated basis (vendor quote / RSMeans Q1 2026 / regional bid spread / engineer estimate)
   • Schedule impact (positive / neutral / negative; days)
   • LCC delta over the owner's hold period (positive / neutral / negative; magnitude if material)
   • Functional / aesthetic trade-off in plain language (one or two sentences the owner can read)
   • Code / warranty / approval risk (AHJ review required, manufacturer warranty affected, spec-section substitution-request required, EOR re-stamp required)
   • Decision owner (owner / architect / EOR / AHJ — name the right authority)
   • Implementation lead time — how long to execute (e.g., 1 week for substitution request; 4 weeks for re-design; 8 weeks for AHJ re-submission)
   • Recommendation (proceed / discuss / hold)

5. Flag the "Do Not VE" set — items that look expensive but should not be VE'd: life-safety, structural, seismic, owner-stated constraints, items already released for fabrication or delivered, items tied to a third-party warranty or commissioning agreement, items where the LCC penalty exceeds the first-cost saving over the hold period.

6. Build the prioritization view:

   • Rank options by savings ÷ implementation effort (Pareto score)
   • Sub-rank by owner-decision burden — options the architect can decide alone are easier to land than options that need owner sign-off in a meeting
   • Total the recommended-bundle savings vs. budget gap; if the bundle does not close the gap, name the next-tier options to consider and the trade-offs they carry
   • Surface a net schedule impact if the bundle is accepted

7. Produce the owner-facing VE workshop materials:

   • One-page summary table (VE #, title, savings range, schedule impact, LCC delta, decision owner, recommendation)
   • Detailed VE-log entries for each option
   • Owner-decision memo: questions to bring to the workshop, decisions needed, deadline for each decision
   • Implementation-path memo for the design team

Output requirements:

Markdown VE log with this structure:

# Value Engineering Log — [Project] — [Phase] — [Date]

**Project:** [Name / Number]
**Phase:** [SD / DD / CD / Post-bid / Mid-build]
**Delivery method:** [DBB / DB / CMAR / IPD]
**Budget gap:** $[X] / [Y%]
**Owner constraints:** [stated constraints]

## Summary Table
| VE# | Title | CSI Div | Savings (low/likely/high) | Schedule | LCC | Decision Owner | Recommendation |
|---|---|---|---|---|---|---|---|

## Detailed VE Log
### VE-001 — [Title]
- **Targets:** [CSI Division / line item]
- **Savings:** $[low] / $[likely] / $[high] — basis: [stated basis]
- **Schedule impact:** [positive / neutral / negative; days]
- **LCC delta over [N]-year hold:** [+/− $X; basis]
- **Trade-off:** [plain-language description]
- **Approval path:** [architect / EOR / AHJ / spec-section substitution request]
- **Implementation lead time:** [duration]
- **Recommendation:** [proceed / discuss / hold]

[...repeat per option...]

## Do Not VE
- [Item] — [reason — life-safety / structural / owner constraint / LCC-negative / already released]

## Recommended Bundle
- VE #s [list] — total savings $[X] (low) to $[Y] (high) — net schedule [days]
- Closes budget gap? [Yes / No — gap remaining $X]
- Next-tier options if gap is not closed: [list]

## Owner Decision Memo
- Questions for owner: [list]
- Decisions needed by: [date — driven by bid validity / GMP reconciliation / NTP target]

---

_VE savings ranges are estimates for discussion. Final savings require sub/vendor quotes, architect substitution-request approval, and EOR/AHJ review where applicable. This VE log was prepared with AI assistance and does not replace a formal VE workshop with the design team._

• Severity color-code is not used here; instead, recommendations are explicit (proceed / discuss / hold)
• Saved to outputs/ if the user confirms

Example Output

Example input:

Project: Brookline MOB TI Phase 2 (the same medical-office TI as the bid-proposal example). Phase: post-bid GMP reconciliation. Delivery: CMAR with GMP. Owner: Brookline Medical Partners LLC, owner-occupier with 15-year hold horizon. GMP came in at $4,797,000 vs. owner budget cap of $4,550,000 — gap is $247,000 / 5.4%. Hard constraints: substantial completion date 2026-12-15 cannot move (lease commitments to incoming tenant practices); LEED is not pursued; ICRA Class III/IV must be maintained; no compromise on med-gas distribution per spec 22 62 13; no compromise on slab-to-deck rated joint detail at demising walls. LCC matters — owner self-occupies and operates. Cost drivers: HVAC distribution at $480k, casework/millwork at $320k, demountable glass partition system at $190k, flooring (carpet tile + porcelain + LVT) at $172k, lighting at $145k, casework countertops solid surface at $48k, painting at $86k.

Expected output:

Value Engineering Log — Brookline MOB TI Phase 2 — Post-Bid GMP Reconciliation — 2026-04-27

Project: Brookline MOB TI Phase 2 / 2026-018 Phase: Post-bid (CMAR GMP reconciliation) Delivery method: CMAR with GMP per AIA A102-2017 Budget gap: $247,000 / 5.4% over owner cap of $4,550,000 Owner constraints: SC date 2026-12-15 cannot move; ICRA Class III/IV maintained; no compromise on med-gas (22 62 13); no compromise on slab-to-deck rated joint detail at demising walls; 15-year hold; LEED not pursued

Summary Table

VE#	Title	CSI Div	Savings (low/likely/high)	Schedule	LCC	Decision Owner	Recommendation
VE-001	Demountable glass partition → conventional gyp w/ vision panels at non-exam offices	10 22 39 / 09 21 16	$48k / $58k / $68k	0 days	Neutral	Owner + Architect	Proceed
VE-002	Solid-surface countertops → laminate at non-exam locations, retain SS at exam	12 36 23	$14k / $18k / $22k	0 days	Slight LCC penalty (~$0.5k/yr maint)	Owner	Discuss
VE-003	Carpet tile manufacturer alternate (Shaw → Mohawk equivalent commercial-grade)	09 68 13	$8k / $11k / $14k	0 days	Neutral	Architect (substitution request)	Proceed
VE-004	Lighting fixture alternates (specified Eaton → approved-equal Lithonia)	26 51 00	$14k / $18k / $22k	+1 wk shorter lead	Slight LCC positive (LED efficacy equal; control compatibility verified)	Architect (substitution request)	Proceed
VE-005	HVAC: keep VAV system family but re-balance VAV-box manufacturer to non-proprietary	23 36 00	$9k / $13k / $16k	Neutral	Neutral	EOR (substitution request, balance report verified)	Proceed
VE-006	Painting: 1 prime + 1 finish in storage / mech rooms only; retain 1+2 in occupied spaces	09 91 23	$5k / $7k / $9k	0 days	Neutral	Architect	Proceed
VE-007	Door hardware grade: ANSI Grade 2 in non-public spaces; retain Grade 1 in main corridor and exam	08 71 00	$4k / $6k / $8k	0 days	Slight LCC penalty (~5-7yr replacement vs. 10-15yr)	Architect	Discuss
VE-008	Casework: laminate veneer in storage and back-of-house only; retain plastic-laminate flush-overlay in exam	12 35 30	$9k / $13k / $17k	0 days	Neutral	Architect	Proceed
VE-009	Defer break-room flooring upgrade (porcelain → LVT until year-2 warranty owner-funded upgrade)	09 65 19	$6k / $9k / $12k	0 days	LCC-negative (porcelain 25yr vs. LVT 10-15yr)	Owner	Discuss — first-cost-positive but LCC-negative on 15-yr hold
VE-010	Defer interior signage to post-occupancy owner-direct contract	10 14 00	$11k / $14k / $17k	0 days	Neutral	Owner	Proceed
VE-011	Reduce CM contingency from 3% to 2.5% per A102 §6.4	n/a	$22k	Neutral	Owner (CM agreement)	Discuss — increases CM exposure; recommend hold
VE-012	Compress GC duration by 1 week via single-night-shift instead of double on rough-in	01 (GCs)	$8k / $11k / $14k	-7 days (favorable)	Neutral	CM (operational)	Proceed

Detailed VE Log

VE-001 — Demountable glass partition → conventional gyp with vision panels at non-exam offices

• Targets: Spec 10 22 39 (demountable glass partitions) at 12 non-exam offices; retain demountable system at 6 exam-room walls
• Savings: $48k / $58k / $68k — basis: vendor quote from KI vs. specified Modernfold; conventional 09 21 16 buildup at $14/SF vs. demountable at $42/SF for the 1,400 SF affected
• Schedule impact: Neutral (gyp framing crew already mobilized; no schedule risk)
• LCC delta over 15-yr hold: Neutral — demountable's LCC value is in reconfigurability; non-exam offices unlikely to be reconfigured during hold
• Trade-off: Loses the demountable system's reconfigure-without-trade-call benefit at non-exam offices. Exam rooms (which actually move per practice growth) retain demountable. Visual continuity maintained at the corridor side via vision-panel sidelights in gyp.
• Approval path: Architect spec-section §3 substitution request; Owner sign-off
• Implementation lead time: 2 weeks (substitution request + owner workshop)
• Recommendation: Proceed — single largest savings in the bundle; LCC-neutral; matches actual reconfigure-likelihood

VE-002 — Solid-surface countertops → laminate at non-exam locations

• Targets: Spec 12 36 23 (solid surface) at break-room and admin-station counters (~80 LF); retain SS at exam-room counters and reception desk for cleanability
• Savings: $14k / $18k / $22k — basis: regional bid spread between Corian-equivalent and high-pressure laminate flush-overlay
• Schedule impact: Neutral
• LCC delta over 15-yr hold: Slight LCC penalty (~$0.5k/year incremental maintenance and edge-replacement on laminate vs. SS); $7.5k over 15-yr
• Trade-off: Laminate requires more careful maintenance and can show wear at edges in 8-10 years; SS has superior cleanability for medical settings (retained at exam rooms). Reception desk remains SS for visual presentation
• Approval path: Architect (in-spec alternate) + Owner (visual sign-off on laminate sample)
• Implementation lead time: 1 week (laminate samples already approved-equal in spec)
• Recommendation: Discuss — first-cost-positive but small LCC penalty; owner's call. Recommend retain at exam, take savings at break-room and admin

VE-003 — Carpet tile manufacturer alternate (Shaw → Mohawk)

• Targets: Spec 09 68 13 — Shaw EcoWorx specified, Mohawk commercial-grade approved-equal alternate
• Savings: $8k / $11k / $14k — basis: factory bid pricing from Mohawk supplier
• Schedule impact: Neutral
• LCC delta over 15-yr hold: Neutral — both products carry equivalent commercial 15-yr warranty
• Trade-off: Architect must approve aesthetic match. Color match to specified pattern verified at sample submittal.
• Approval path: Architect via CSI 13.1 substitution request
• Implementation lead time: 1 week
• Recommendation: Proceed

VE-004 — Lighting fixture alternates (Eaton → Lithonia approved-equal)

• Targets: Spec 26 51 00 — proprietary Eaton specified
• Savings: $14k / $18k / $22k — basis: distributor quote, 320 fixtures
• Schedule impact: +1 week shorter lead (Lithonia 4-wk vs. Eaton 6-wk lead at current 2026 supply)
• LCC delta over 15-yr hold: Slight LCC positive — equivalent LED efficacy per IES report; verified compatibility with specified DALI controls; no driver-replacement burden differential
• Trade-off: Aesthetic equivalence verified by architect. Driver compatibility verified.
• Approval path: Architect via CSI 13.1 substitution request
• Implementation lead time: 1 week
• Recommendation: Proceed — first-cost-positive, LCC-neutral-or-positive, schedule-positive

VE-009 — Defer break-room flooring upgrade

• Targets: Spec 09 65 19 — porcelain in break room replaced with LVT
• Savings: $6k / $9k / $12k
• Schedule impact: Neutral
• LCC delta over 15-yr hold: LCC-negative — porcelain has 25-year service life and zero replacement; LVT in a high-traffic break room has 10-15-year service life; LCC penalty $4-8k over the hold period likely exceeds the first-cost saving
• Trade-off: Owner saves first cost; pays it back (or more) in mid-hold replacement
• Approval path: Architect; Owner
• Implementation lead time: 1 week
• Recommendation: Discuss — flagged as LCC-negative on 15-yr hold. Owner may still accept if budget gap is the binding constraint

VE-011 — Reduce CM contingency from 3% to 2.5%

• Targets: A102 §6.4 CM-controlled contingency
• Savings: $22k (mechanical reduction; not a scope change)
• Schedule impact: Neutral
• LCC delta: Neutral
• Trade-off: Increases CM exposure on unforeseen conditions. At 2.5% on a $4.4M cost-of-the-work, contingency is $109k, which is at the low end for a TI of this complexity (occupied adjacencies; med-gas; rated demising)
• Approval path: A102 amendment; Owner + CM
• Implementation lead time: 1 day (contract amendment)
• Recommendation: Discuss — recommend hold. This is not VE; it is risk transfer. If accepted, recommend the owner accept a corresponding GC-fee bump to offset the carried risk

Do Not VE

• Slab-to-deck rated joint detail at demising walls — owner stated constraint; also a code condition under UL U419 listing
• Med-gas distribution (22 62 13) — owner stated constraint; also code-conditioned per NFPA 99
• ICRA / ILSM (interim life safety measures) — required to maintain Class III/IV protocols during construction; not VE territory
• Substantial completion date — owner stated constraint driven by lease commitments to incoming tenants
• Pre-bought med-gas distribution components — already released for fabrication
• Builders' Risk insurance — required by lender; not VE territory
• Performance and payment bonds — required by AIA A102 contract terms

Recommended Bundle

Recommended (Proceed): VE-001, VE-003, VE-004, VE-005, VE-006, VE-008, VE-010, VE-012 Total savings: $114k (low) / $147k (likely) / $182k (high) Net schedule impact: -7 days favorable (VE-012 compresses; VE-004 lead shortened)

Adding owner-discussed (Discuss → assume accepted): VE-002, VE-007 Total bundled savings: $132k (low) / $171k (likely) / $212k (high)

Closes the $247k gap? No, not in the recommended-or-discussed set ($171k likely).

Next-tier options to close the remaining $76k gap:

• VE-009 (break-room flooring) — $9k likely; LCC-negative — owner accepts the LCC trade
• VE-011 (CM contingency reduction) — $22k; recommend hold; if owner pushes, recommend offsetting GC-fee bump
• Beyond the log: Re-open a structural-grid VE pass for the demising walls (currently slab-to-deck full UL U419 — could portions revert to plenum-only ratings if AHJ accepts? requires AHJ pre-approval and is high-risk for the schedule). Or owner accepts a $76k gap and absorbs into owner contingency (5% / $228k available).

Recommended path: take the $171k likely bundle; absorb the remaining $76k into owner contingency with a 90-day buyout-watchlist commitment from CM to find additional savings during sub buyout (per config.yml buyout-watchlist-posture default).

Owner Decision Memo

Questions for owner workshop (target date: 2026-05-05):

1. VE-001 (largest savings): accept demountable-to-gyp at non-exam offices? Architect supportive on visual-continuity grounds.
2. VE-002, VE-007 (small LCC penalties): accept LCC trade for first-cost saving? Owner-occupier with 15-year hold — these add up to ~$15k LCC penalty against $24k likely first-cost saving on a 15-year basis (roughly net-positive but tight).
3. VE-009 (LCC-negative): owner's call on whether budget gap pressure overrides 15-year LCC.
4. VE-011 (contingency reduction): owner's call on absorbing CM-side risk via reduced contingency. Not recommended unless paired with offsetting fee adjustment.
5. Remaining gap acceptance ($76k after recommended bundle): owner-side absorption from owner contingency vs. additional VE rounds?

Decision deadline: 2026-05-05 (NTP scheduled 2026-06-15; substitution requests need 2 weeks plus architect review, so VE decisions must close 5 weeks before NTP to avoid schedule slip).

---

VE savings ranges are estimates for discussion. Final savings require sub/vendor quotes, architect substitution-request approval, and EOR/AHJ review where applicable. This VE log was prepared with AI assistance and does not replace a formal VE workshop with the design team. CM contingency reduction is a risk-transfer item, not VE; recommend addressing separately from the VE log if pursued.

Reviewer-of-Platform-AI-Output Sub-Mode

Construction platforms in 2026 increasingly produce AI-generated VE candidate lists from the same estimate / spec / model the GC is already working in: Procore Datagrid AI VE agent (summer-2026 broader-availability trajectory; produces VE candidates with savings ranges from line-item estimate + spec parse), Beam AI cost-reduction recommendations (estimating-platform-side VE list from the takeoff + estimate), ALICE Technologies optioneering (schedule-VE: phasing / sequencing alternates that compress GC duration — McKinsey partnership 35+ clients), Trunk Tools VE workflow (spec-substitution candidates from the spec parse), Document Crunch / Trimble Construction One (post-acquisition: spec relaxation candidates from CDs + AHJ amendments), Autodesk Construction Cloud + AI Assistant (spec + drawing-parse VE candidates), Togal.AI (quantity / coverage reduction candidates from the AI takeoff), STACK (assembly-substitution VE candidates), and general LLM workflows where the chief estimator pastes the estimate into ChatGPT / Claude and accepts a VE candidate list. The chief estimator / preconstruction lead is then the reviewer-of-AI-output. This sub-mode produces a redline of the platform's VE candidate list before it goes into the VE workshop materials.

When the input is a platform-AI VE list (not a raw estimate), apply this six-point redline check before accepting candidates into the recommended bundle:

1. Phase-altitude alignment. Platforms produce VE candidates without checking whether the altitude matches the project phase. A CD-phase project that gets SD-altitude candidates (system change: VRF → packaged rooftop) will mostly reject those candidates; the redline downsizes to phase-appropriate altitude per the Phase / VE-able table above. Flag and drop any candidate whose altitude is one or more tiers above the current phase (e.g., system change proposed at CD; concept-level shape change proposed at DD).
2. Life-safety / structural / seismic exclusion check. Platforms occasionally propose VE candidates that touch life-safety, structural, seismic, smoke evacuation, or egress (e.g., "reduce fire-rating from 2-hr to 1-hr at corridor partition," "reduce shear-wall thickness," "substitute fire-sprinkler heads to less-expensive model without re-running hydraulics"). The redline drops these unconditionally and surfaces them in the "Do Not VE" section. The hard rule applies — no exceptions without EOR review and AHJ re-submission flagged as a required step.
3. Owner-constraint compliance. Platforms do not read the owner's stated constraints (LEED level, occupancy date, accessibility commitments, brand-specific finishes the owner chose, M/W/DBE participation goal). Cross-check every platform candidate against the owner's stated constraints from the input and drop any candidate that breaches a stated constraint. Flag explicitly so the workshop materials show what the platform missed.
4. LCC-impact disclosure (the v2.0 seventh category, enforced here). Platforms typically report first-cost savings without computing the LCC delta over the owner's hold period. For owner-occupier and institutional-owner projects (per config.yml LCC-discipline posture), compute the 20-year (or owner-hold-period) LCC delta for every platform candidate and flag any that are first-cost-positive / LCC-negative (e.g., porcelain → LVT in a high-traffic break room over a 15-year hold; lower-SEER HVAC over a 20-year hold; lower-efficacy LED over a 10-year hold). For developer-flip projects, this check is neutralized per config.yml posture.
5. Savings-basis credibility. Platforms generate savings ranges with named or unnamed bases. The redline reads each candidate's basis and rates it H / M / L credibility: H = vendor-specific quote or regional bid-spread within 90 days; M = RSMeans / national-database reference within the current quarter; L = engineer estimate / "industry typical" with no anchor. Any candidate with L credibility is held until a vendor quote or sub-of-record confirmation is in hand; the workshop materials present these as "candidate — pending verification" not "recommended — proceed."
6. Approval-path completeness. Platforms often surface the savings without the implementation path (architect substitution-request / EOR re-stamp / AHJ re-submission / spec-section approved-equal). The redline fills in the missing approval path per the spec section, the project's delivery method (DBB / DB / CMAR / IPD changes who has authority), and the implementation lead time (1 wk substitution request / 4 wk re-design / 8 wk AHJ re-submission). Any candidate whose approval-path lead time exceeds the available pre-NTP window is dropped to "next-tier" rather than "recommended."

Sub-mode output: (a) platform's original AI-generated VE candidate list (preserved verbatim, including savings ranges and any narrative the platform produced), (b) redline applied to each of the six points with phase-altitude downsizing / hard-rule drops / owner-constraint drops / LCC-impact annotations / credibility ratings / restored approval paths, (c) final estimator-accepted VE log ready for the workshop, with pre-redline-vs-post-redline candidate-count delta (e.g., platform proposed 24 candidates → 18 accepted-as-is + 4 downsized-by-phase + 2 dropped-on-owner-constraint), (d) provenance footer noting which platform produced the list, which version of its AI (if known), what the redline changed, what the estimator accepted, and a pattern-notes feedback-loop for vendor-config calibration over time (e.g., "platform consistently proposes structural VE on healthcare projects — adjust prompt to exclude structural Pareto items upstream").

Example Output — Reviewer-of-Platform-AI-Output Sub-Mode (Procore Datagrid AI VE Agent)

Example input scenario: Same Brookline MOB TI Phase 2 post-bid GMP-reconciliation as above ($247K gap on $4,797K GMP, 15-yr owner-occupier hold, ICRA Class III/IV maintained, med-gas + slab-to-deck rated joint detail off-limits). The Stonebridge preconstruction lead ran the GMP-reconciliation estimate through Procore Datagrid AI's VE candidate agent (private-beta cohort access). Datagrid produced an 18-candidate VE list:

Procore Datagrid AI VE Candidates — Brookline MOB TI Phase 2

1. Demountable glass partitions → conventional gyp throughout (all 18 walls): $74K
2. Solid-surface countertops → laminate throughout (incl. exam): $24K
3. Carpet tile alternate manufacturer (Shaw → Mohawk): $11K
4. Lighting alternate (Eaton → Lithonia approved-equal): $18K
5. VAV system → packaged rooftop (system change): $89K
6. Reduce demising wall rating from 2-hr to 1-hr at non-corridor locations: $34K
7. Reduce med-gas spec from oxygen-cleaned copper to general copper: $42K
8. Painting 1 prime + 1 finish in all spaces: $14K
9. Door hardware Grade 1 → Grade 2 throughout (incl. main corridor): $19K
10. Casework laminate veneer throughout (incl. exam): $28K
11. Defer break-room flooring porcelain → LVT: $9K
12. Defer interior signage: $14K
13. Reduce CM contingency 3% → 2%: $44K
14. Reduce GC contingency 5% → 3%: $91K
15. Slab-on-grade thickness 5" → 4": $11K
16. Reduce LEED documentation effort (project is not LEED): $3K
17. Compress GC duration 1 wk via single night shift: $11K
18. Substitute fire-sprinkler heads to less-expensive model: $9K

The preconstruction lead is the reviewer-of-AI-output. Apply the six-point redline before accepting into the workshop bundle.

Expected output:

VE PLATFORM-AI OUTPUT REVIEW — REDLINE

Platform: Procore Datagrid AI VE agent (private-beta cohort, May 2026) Subject: Brookline MOB TI Phase 2 — VE candidate list — Post-bid GMP reconciliation Reviewed by: Preconstruction Lead (AI-assisted), Stonebridge Construction Review Date: 2026-04-27 Project posture per config.yml: owner-occupier 15-yr hold = LCC-discipline ON; CM-contingency policy = don't-VE-contingency-without-fee-offset; phase = post-bid GMP reconciliation (system change is mostly out of altitude)

PLATFORM VE CANDIDATE LIST (PRESERVED VERBATIM)

18 candidates as listed by Procore Datagrid AI, $574K nominal total at the high end if all accepted.

SIX-POINT REDLINE

1. Phase-Altitude Alignment — 1 candidate downsized, 1 dropped.

• Candidate 5 (VAV → packaged rooftop) is a system change — SD altitude, proposed at post-bid GMP reconciliation phase. Dropped to "next-tier — would require re-design 8 weeks + EOR re-stamp + AHJ re-submission; not feasible inside NTP window." Surface in the workshop materials as "platform-proposed, infeasible at this phase" with the rationale.
• Candidate 17 (compress GC duration via single night shift) — sequencing / productivity altitude, in-phase. Accept.

2. Life-Safety / Structural / Seismic Exclusion — 3 candidates dropped to "Do Not VE."

• Candidate 6 (reduce demising wall rating from 2-hr to 1-hr at non-corridor locations) — fire-rating reduction; UL U419 listing condition; owner-stated constraint also explicit ("no compromise on slab-to-deck rated joint detail at demising walls"). Dropped — Do Not VE. Hard rule + owner constraint.
• Candidate 7 (reduce med-gas spec from oxygen-cleaned copper to general copper) — NFPA 99 condition; owner-stated constraint also explicit. Dropped — Do Not VE. Hard rule + owner constraint + code.
• Candidate 18 (substitute fire-sprinkler heads to less-expensive model) — fire-suppression system; hard rule says no without EOR review + AHJ re-submission; hydraulic-calc impact not evaluated by platform. Dropped — Do Not VE. Hard rule.
• Candidate 15 (slab-on-grade thickness 5" → 4") — structural; outside ICRA Class III/IV demolition / repour during occupied operations is not viable; even if structurally OK, the candidate touches the structural design. Dropped — Do Not VE. Hard rule.

3. Owner-Constraint Compliance — 2 candidates dropped (overlapping with #2).

• Already captured in #2: candidates 6 + 7 breach explicit owner constraints. No additional candidate drops at this point.
• Candidate 16 (reduce LEED documentation effort) — project is not LEED per the input; the platform proposed a candidate that does not apply. Dropped — does-not-apply (no actual savings to capture).

4. LCC-Impact Disclosure — 4 candidates LCC-annotated.

• Candidate 1 (demountable → gyp, all 18 walls): platform proposed all walls. The v2.0 example carved 12 non-exam + 6 exam (retain demountable at exam for reconfigurability over 15-yr hold). The platform's full-stripout would deprive owner of $42K NPV reconfigurability benefit at exam over 15-yr hold. Downsize to non-exam only (12 walls) — accept $58K of the $74K platform claim; LCC-aligned.
• Candidate 2 (solid-surface → laminate throughout, incl. exam): exam-room counters drive cleanability; LCC penalty of laminate edge-wear over 15-yr at exam stations exceeds the first-cost saving. Downsize to non-exam only — accept $18K of the $24K platform claim; LCC-aligned.
• Candidate 10 (casework laminate veneer throughout, incl. exam): exam casework drives medical-cleanability; same LCC reasoning. Downsize to non-exam only — accept $13K of the $28K platform claim; LCC-aligned.
• Candidate 11 (defer break-room flooring porcelain → LVT): flag as first-cost-positive / LCC-negative over 15-yr hold (porcelain 25-yr vs. LVT 10-15-yr; $4-8K LCC penalty against $9K likely first-cost saving). Surface to owner with the LCC framing; not auto-accepted.

5. Savings-Basis Credibility — 3 candidates dropped or held.

• Candidate 13 (CM contingency 3% → 2%, $44K) — not VE; risk transfer. Per config.yml CM-contingency policy "don't-VE-contingency-without-fee-offset," hold and surface as risk-transfer-not-VE to the owner with the recommended GC-fee bump offset.
• Candidate 14 (GC contingency 5% → 3%, $91K) — not VE; risk transfer; same posture. Hold; not recommended at the workshop without offsetting risk-allocation discussion. Additionally, dropping owner's contingency from 5% to 3% on a TI of this complexity (occupied healthcare adjacencies; ICRA discipline; med-gas; rated demising) leaves only $137K cushion on a $4.55M cap — historically tight.
• Candidate 9 (door hardware Grade 1 → Grade 2 throughout, including main corridor) — main corridor hardware is the spec floor (Grade 1 in main corridor + exam; Grade 2 acceptable in back-of-house). Platform over-reached. Downsize to non-corridor only — accept $6K of the $19K platform claim.

6. Approval-Path Completeness — Restored on accepted candidates.

• Candidates 1 (demountable substitution), 2 (countertop substitution), 3 (carpet manufacturer), 4 (lighting alternate), 5 (already dropped), 9 (door hardware grade), 10 (casework veneer): architect via CSI 13.1 substitution request; 1–2 weeks each.
• Candidate 8 (painting 1 prime + 1 finish in mech / storage only — downsize from "all spaces"; retain 1+2 in occupied per spec): architect + owner sign-off on sample; 1 week.
• Candidate 12 (defer signage): owner-direct contract post-occupancy; 1 day.
• Candidate 17 (compress GC duration via shift change): CM operational decision; 1 day to commit.

ACCEPTED VE LOG (POST-REDLINE)

VE#	Title	Savings (low/likely/high)	Notes
VE-001	Demountable → gyp at 12 non-exam offices (downsized from platform's "all 18 walls")	$48k / $58k / $68k	LCC-aligned per redline
VE-002	Solid-surface → laminate at non-exam only (downsized)	$14k / $18k / $22k	LCC-aligned per redline
VE-003	Carpet tile alternate (Shaw → Mohawk)	$8k / $11k / $14k	Accept as-platform
VE-004	Lighting alternate (Eaton → Lithonia)	$14k / $18k / $22k	Accept; +1wk lead shorter
VE-006	Painting 1 prime + 1 finish in mech / storage only (downsized from "all spaces")	$5k / $7k / $9k	Retain 1+2 in occupied per spec
VE-007	Door hardware Grade 1 → Grade 2 at non-corridor only (downsized)	$4k / $6k / $8k	Main corridor + exam retain Grade 1
VE-008	Casework laminate veneer at non-exam only (downsized)	$9k / $13k / $17k	LCC-aligned
VE-009	Defer break-room porcelain → LVT (flagged LCC-negative)	$6k / $9k / $12k	Surface to owner with LCC framing
VE-010	Defer interior signage (post-occupancy)	$11k / $14k / $17k	Accept
VE-012	Compress GC duration 1 wk via single night shift	$8k / $11k / $14k	Accept; -7 days schedule

Total likely savings (post-redline): $165K (vs. platform's $574K nominal max) — closes ~67% of $247K gap with the recommended-or-flagged set. Pair with owner contingency for the residual ~$82K + 90-day buyout watchlist commitment per config.yml.

DO NOT VE (RESTORED FROM REDLINE)

• Candidate 5 — VAV → packaged rooftop: dropped on phase-altitude (SD altitude proposed at post-bid; 8-wk re-design + EOR + AHJ infeasible inside NTP window)
• Candidate 6 — demising wall rating reduction: owner constraint + UL listing + code
• Candidate 7 — med-gas spec relaxation: owner constraint + NFPA 99
• Candidate 15 — SOG thickness reduction: structural + ICRA occupied-operations re-pour infeasibility
• Candidate 16 — LEED documentation reduction: does not apply (project is not LEED)
• Candidate 18 — sprinkler-head substitution: life-safety; hard rule; hydraulic re-calc impact not evaluated

HOLD — RISK TRANSFER (NOT VE)

• Candidate 13 — CM contingency 3% → 2%: $44K savings; surface as risk-transfer with recommended GC-fee bump offset
• Candidate 14 — GC contingency 5% → 3%: $91K savings; not recommended at this complexity

PROVENANCE + PATTERN NOTES

• Platform: Procore Datagrid AI VE agent (private-beta cohort, May 2026)
• Pre-redline candidate count: 18
• Accepted (recommended or discuss): 10
• Downsized (LCC or scope): 5 (candidates 1, 2, 8, 9, 10)
• Dropped to Do Not VE: 6 (candidates 5, 6, 7, 15, 16, 18)
• Held — risk transfer (not VE): 2 (candidates 13, 14)
• Pattern notes for Procore Datagrid AI VE feedback-loop:
  • Datagrid consistently proposes structural / fire-rating VE on healthcare projects — feed back the rule "exclude life-safety, structural, seismic, fire-rating candidates" upstream
  • Datagrid did not weight the owner's 15-yr hold against LCC — feed back the owner-hold-period parameter so LCC-negative candidates get auto-annotated
  • Datagrid did not respect the owner-stated-constraint list (med-gas, rated joint) — feed back the constraint list as input filter
  • Datagrid proposed contingency reduction as VE — re-label as risk-transfer-not-VE
  • Datagrid over-applied substitutions to all spaces (demountable / hardware / casework / countertops) when the right answer was non-exam / non-corridor only — feed back the "downsize-by-zone" rule
• Disclaimer: This redline is AI-assisted. VE candidates require formal architect substitution-request approval, EOR re-stamp where applicable, and AHJ re-submission where applicable. The platform's first-pass savings reduction (~$574K → $165K accepted) reflects the LCC-discipline, hard-rule, and owner-constraint filters that distinguish a VE workshop from a candidate-generation exercise.