Two Known Markets.
One Unknown Market.
One Platform.
ML Systems operates across three compounding markets totaling $63 trillion — larger than the US national debt. Construction. Banking. And a deconstruction economy that doesn't exist yet.
Total Addressable Market
$63 Trillion. Larger Than the US National Debt.
Two well-understood markets. One we're creating from scratch.
Known Market
$36T
Construction
The asset we rebuild. US construction spending exceeds $2T/year; the global stock of built environment represents a $36T addressable market for the deconstruct-and-rebuild cycle.
Known Market
$13.5T
Banking / Mortgage
The product we redesign. The US residential mortgage market alone is $13.5T outstanding. The Reversed Conventional Mortgage creates a new origination product inside this existing market.
Unknown Market
$13.5T
Deconstruction
The market we create. Material recovery at 80–90% converts demolition waste into a resale commodity. No scaled operator owns this yet — ML Systems is building the category.
$63T combined — 1.75× the US national debt (~$36T)
One integrated platformWhy ML Systems Wins
Four Structural Advantages. Each Compounds the Last.
L1. — RCM
Reversed Conventional Mortgage
Recurring interest income + equity loop retention
100% of each payment hits principal first. Interest accrues as a separate liability. Clients build equity faster and loop back into the next project — generating origination revenue local competitors never had.
L2. — Deconstruction
80–90% Material Recovery
~4–6 ppt gross margin lift vs. baseline
Our 2-day crane sequence recovers the structural material. 51% is resold, reused, or recycled. The remainder offsets material purchases on the next build — converting cost into revenue.
L3. — Robotization
Tier 1 Live 2026 · Robotic Integrations Planned 2027
$57,600/year per home in labor savings
Tier 1 robots handle material staging, transport, and movement. Full robotic integrations — Tier 2 partial carpenter displacement and electric on-site movement — are planned for 2027, adding another $3–4K/month in savings.
L4. — ML Hybrid
Highly Unique Composite Structure
Over-engineered for multi-cycle recovery · ~$15K/home material reduction
Every ML Hybrid home is Highly Unique — shaped by the owner's five-step curriculum choices. Concrete + steel composite vs. conventional wood frame, over-engineered for multi-cycle recovery. Structural moat, not procurement advantage.
Margin Architecture
Local Competition 16% → ML Systems 24–27%
Every margin layer is grounded in real operational data from 269 ML1 work days on local Rhode Island job sites.
Local Competition Baseline
269 ML1 work days, RI job sites, 2020–2022
Material Recovery
80–90% recovery; 51% resale/reuse reduces material purchases by ~$50K/home
Robot Tier 1 (2026)
Laborer displacement — $5,636/month fully loaded; $57,600/year per home
Robot Tier 2 (2027+)
Partial carpenter displacement — est. $3,000–$4,000/month additional
Fuel Elimination
Electric robot movement replaces ~$2,000–$4,000/month fuel opex (partial)
Scale Procurement
Volume purchasing leverage on materials as portfolio scales past 10 homes/year
ML3 Scoring Uplift
AI efficiency score (0–100) optimizes job sequencing and reduces rework cost
ML Hybrid (structural system)
Concrete + steel composite vs. conventional wood frame — over-engineered for multi-cycle recovery; est. ~$15K/home material reduction at scale
Benchmark: Toll Brothers FY2024: 26% gross margin on luxury residential
24–27% targetPath to $500M
Conservative. Grounded. Compounding.
10-year model built from local competition unit economics — same crew structure, same RI market, conservative growth assumptions.
| Year | Homes | Revenue | Gross Profit | Net Income |
|---|---|---|---|---|
| 2026 | 1 | $1.5M | $345K | $120K |
| 2027 | 2.2 | $3.3M | $759K | $264K |
| 2028 | 4 | $6.0M | $1.4M | $480K |
| 2029 | 8 | $12.0M | $2.8M | $960K |
| 2030 | 20 | $30.0M | $6.9M | $2.4M |
| 2035 | 270 | $500M | $115M | $40M |
AI integration (ML3 full autonomy, ontology licensing) not modeled — represents additional upside.
The Investment
18-month runway · AI-powered platform · Full value chain
$1,692,146
S(9,131 days · 25 yrs) = $41,692,146 − $40,000,000 = $1,692,146
Primary Use
Crane Services
The 2-day crane sequence is the operational core of ML Systems — it enables 80–90% material recovery and converts demolition waste into a resale commodity. Heavy liability exposure (equipment, operator, site) makes crane access the highest-friction barrier to entry in residential deconstruction. This raise is sized around that reality. Funding crane operations is not a cost — it is the competitive moat.
Use of Funds
Capital Assets
Crane equipment + Brokk remote-controlled demolition robot + heavy liability coverage, operator certification, rigging — the deconstruction unlock
Platform & Tech
Portal, mobile app, AI scorer, ML3 infrastructure
Working Capital
First home acquisition, materials, crew ramp, Robot Tier 1 deployment
Operating Reserve
18-month runway buffer — compliance, legal, admin
Return Profile
25.87%
IRR floor — field-validated baseline
25.87% IRR floor validated by local competition 10% net margin on $78,206 initial raise.
18-month runway covers first home completion, full loop validation, and AI scorer deployment.
ML3 AI scoring (0–100) optimizes every build — efficiency gains compound across every cycle.
Platform, mobile app, and autonomous scoring infrastructure built and live on day one.
Repayment · Daily · Arithmetic · 25 years
Founded By
Sal Parvez
Language Modeler / Financial Architect
URI Dean's List — Finance & Management Information Systems. Estimator at Shawnlee Construction. Laborer at a local RI construction firm (269 ML1 work days, 2020–2022). Financial analyst at Leshinsky & Associates.
The ML Systems investment thesis is not theoretical — it is built from direct field experience on RI job sites, first-hand exposure to a profitable construction company's financials, and a decade of financial modeling practice.
GC registration in progress — Rhode Island
Platform live: mlsystemsri.com + app.mlsystemsri.com
Domains secured across .com, .net, .info, .store, .xyz
ML1 dataset: 269 days, 1,480 task executions — robot training ready
ML1 Dataset
The only construction robotics dataset built by a founder performing real work.
269 work days. 1,480 task executions. 81 unique task codes. 32 atomic robot actions. Every entry recorded from the field — not extrapolated from third-party data.
269
Work Days
1,480
Task Executions
81
Unique Task Codes
32
Atomic Robot Actions
Ready to Invest in What's Next?
Request the full pitch deck or schedule a conversation.
Get in Touch