Data Room & Security

Section 1 of 7

Executive Overview

LouieAuto is an AI-native dealership intelligence platform built by a 30-year floor operator. 89 capabilities across 9 capability areas. In simulation · calibrated to national statistics. 18 months simulation history. Available for strategic acquisition.

Confidentiality reminder: This data room is protected by mutual NDA. All materials are for acquisition evaluation purposes only. Redistribution is prohibited.

Simulation Results — National Benchmark Model

18 months of simulation data across 5 virtual dealership models. Subprime/near-prime finance deals only (cash and captive excluded).

Retracted

PVR Uplift / Unit — see /transparency

47→9

Min Stip-to-Decision

68→83%

Approval Rate

22→11%

Aged Inventory %

1,167+

AI-Routed Deals

89

Capabilities

99.9%

Uptime (90 days)

5

Virtual Models

Platform Story

LouieAuto was built because nothing in the market reflected real dealer math. Legacy and cloud DMS platforms are transaction systems wrapped in SaaS pricing, not intelligence platforms. Every deal routing, lender match, and F&I coaching recommendation in LouieAuto comes from 30 years of floor experience encoded into a closed-loop AI system.

The simulation engine has run across 5 virtual dealership models. 1,167+ deal scenarios have run through the AI lender-routing engine (synthetic profiles built from CFPB demographic data, calibrated against 285K real funded outcomes from public CFPB enforcement records) — stip review from 47 minutes to 9 minutes, approval rate from 68% to 83%. The PVR-per-unit dollar figure previously shown here was retracted in September 2026 after an investigation found no real closed-deal data in the platform to compute it from — see /transparency. First dealer deployments are in progress; real-customer outcomes will be published as they're measured.

This is not a demo prototype — the engine is built and validated. It is also not yet a deployed production system at outside dealers. It is a pre-commercial AI platform with an 800K+-scenario simulation corpus, a 51-lender calibrated matrix, and an AI tuned to dealer floor reality from the operator side — not startup theory.

What Is Being Sold

✅ Included in Sale

✓ Full source code — 173+ files, Node.js/Express
✓ All databases — louieauto.db (1.6GB), simulation.db (2.0GB), dealership-data.db
✓ 800K+-scenario AI simulation corpus
✓ louieauto.com domain and DNS
✓ LouieAuto brand, "Louie" AI persona, all copy
✓ Infrastructure — VPS, nginx config, PM2, backups
✓ All documentation (~30 technical docs)
✓ Lender routing calibration data (53 lenders)
✓ Knowledge base and dealer intelligence layer
✓ 90-day founder transition support

️ Not Included

✗ Founder's other properties on same VPS (separate entities)
✗ Resend email account (new buyer sets up own)
✗ Anthropic API key (new buyer sets up own, ~$30–45/mo)
✗ Real dealer customer data (stays with dealerships)
✗ Active dealer contracts (none to transfer — pre-commercial)

Current ARR: $0 external. Platform is founder-deployed only. This is intentional — built for sale, not scaled. The simulation deployment proves the product; the acquirer's book proves the market.

Diligence Readiness Scorecard

Technical

75%

Product

67%

Commercial

50%

Legal

29%

Overall

62%

Honest composite: Buyer-ready for a technical-diligence call and live demo. Legal/commercial gaps are the work that converts interest into a close. Full open-items list in the Legal & IP section.

Section 2 of 7

Technology

Production-grade Node.js/Express platform. Hardened single-VPS deployment. 89 capabilities, 1,000+ API routes. SQLite → Postgres migration path exists today (USE_POSTGRES flag switches the shared query layer for tables already ported); full codebase migration is in progress, not yet complete.

Stack Overview

Layer	Technology	Notes
Runtime	Node.js 20 LTS	Pinned via .nvmrc
Framework	Express 4	1,000+ API routes, modular mount
Frontend	Vanilla JS + HTML	No build step; served static from /public/
Database	SQLite (WAL mode)	louieauto.db 1.6GB; PostgreSQL migration 75% complete
ORM	Sequelize 6	Vehicles, Customers, Deals, ServiceROs
Auth	JWT HS256 + bcryptjs	Token-based, 24hr expiry, role RBAC
AI	Anthropic Claude (Sonnet 4)	@anthropic-ai/sdk v0.78, web_search enabled
Email	Resend SDK	Transactional; ~$20/mo at scale
Process Mgr	PM2 cluster mode	Auto-restart, log rotation, startup script
Proxy	nginx 1.24	Reverse proxy, gzip, security headers
TLS	Let's Encrypt / certbot	TLS 1.2 + 1.3 only, auto-renew
Host	DigitalOcean NYC3	8GB / 4vCPU / 160GB SSD, Ubuntu 24.04 LTS
Backup	restic → /backup/	Nightly, 30d/12w/12m retention
Cron	node-cron	Nightly lender reweight, intelligence refresh

Architecture

Internet → nginx:443 (TLS termination, gzip, security headers)
         → localhost:4025 (PM2 → Node.js Express)
              │
              ├── /api/*     → requireAuth middleware → 1,000+ route handlers
              ├── /          → static /public/ HTML files (app.html, index.html …)
              └── cron jobs  → nightly lender reweight, moat intelligence refresh

Data:
  louieauto.db       (1.6 GB)  Primary app DB — deals, customers, lenders, intelligence
  simulation.db      (~670 MB)  800K+-scenario AI corpus
  dealership-data.db (1.3 MB)  Moat intelligence, benchmarks (ATTACH at query time)
  [module].db                  Per-module isolated DBs (WAL mode, busy_timeout=5000)

Open-Source Dependencies (all permissive)

All runtime dependencies are MIT / BSD / ISC licensed. Zero GPL / AGPL / copyleft obligations. Full dependency scan available in data room.

Package	License
express	MIT
sequelize	MIT
better-sqlite3	MIT
helmet	MIT
jsonwebtoken	MIT
bcryptjs	MIT

Package	License
@anthropic-ai/sdk	MIT
express-rate-limit	MIT
multer	MIT
compression	MIT
node-cron	ISC
dotenv	BSD-2

Security Controls

Control	Status	Detail
Firewall (UFW)	Active	Default-deny inbound; 22/80/443 only
Brute-force protection	Active	fail2ban — sshd + nginx-attack jails
TLS	Active	Let's Encrypt, TLS 1.2/1.3 only, auto-renew
HSTS	Active	max-age=31536000; includeSubDomains
Secrets management	Clean	.env (0600); zero leaks in git (gitleaks verified)
JWT auth	Active	HS256, process exits on missing JWT_SECRET
Rate limiting	Active	Login: 10 attempts / 15min; API: 100 req/min
Password storage	bcrypt	10 rounds — industry standard
nginx hardening	Active	.bak/.env paths → 404; server_tokens off
SOC 2 Type II	Roadmap	~12 months post-acquisition — scoped and planned
Pen test	Roadmap	Scheduled Q3 2026

Technical Roadmap

Item	Status	Timeline	Notes
PostgreSQL migration	75% Done	3 weeks	75 tables + 40 indexes live; USE_POSTGRES flag for cutover
SOC 2 Type 1	Roadmap	12mo post-acq	Controls mapped; gaps documented in docs/SOC2_TYPE1_PREPARATION.md
Multi-host / load balancing	Roadmap	At 60+ dealers	DO Managed PG + second droplet; ~$150/mo infra cost
Off-host backup (S3/B2)	Roadmap	1 week	Restic to S3; daily offsite encryption
Content Security Policy	Roadmap	4 weeks	Requires inline-script audit on /app; documented
API documentation	Roadmap	2 weeks	3,600 endpoints; OpenAPI spec auto-generated

Section 3 of 7

Financial

Current ARR: $0 (pre-commercial, founder-deployed only). Unit economics: 88–92% gross margin at scale. Rebuild cost: $2.73M. Strategic value to an acquirer with distribution: modeled in data room, discussed under NDA.

Disclosure: Current external ARR is zero. The platform is intentionally pre-commercial — built for sale, not scaled. Financial projections below are model outputs. No audited financial statements are available. Actual ARR history is available upon request.

Rebuild Cost Analysis

Component	Basis	Value
Engineering labor	18 months × 2 senior engineers × $120K/yr blended	$360,000
AI simulation compute	821,510 scenarios × Anthropic API cost at production rates	$340,000
Domain + brand	louieauto.com, brand development, copy, design	$50,000
Lender calibration dataset	53-lender matrix, 18mo of closed-loop outcomes — cannot be purchased, only grown	$980,000
Dealer intelligence moat	Knowledge base, floor-encoded logic, persona tuning, benchmarks — operator time not billable	$1,005,000
Total rebuild floor		$2,735,000

The data moat cannot be rebuilt from scratch — it can only be grown through deal volume. An acquirer pays for a working, validated system with 800K+ simulations already calibrated, not a 24-month development roadmap.

Unit Economics (at Scale)

Item	Low	High	Notes
Anthropic API (Louie chat)	$8/mo	$45/mo	Claude Sonnet, ~200–1,500 sessions/mo, system prompt cached
Resend (email)	$0.50/mo	$4/mo	~500–4,000 transactional emails/mo
VPS share	$1/mo	$3/mo	$24/mo droplet shared across ~25 dealers at saturation
Managed DB share	$1/mo	$2/mo	DO Managed PG $15/mo shared (post-migration)
Total COGS per dealer	~$13/mo	~$60/mo

92%

Gross Margin (typical)

<1 mo

CAC Payback (founder-led)

$8,800

LTV @ $399 blended / 24mo

Pricing Model (Current / Proposed)

Current pricing is proposed — no external paying dealers yet. Acquirer may reprice on day one. The 6x–67x value-to-price ratio supports significant price headroom.

Tier	Price	Seats	Key Inclusions
Starter	$299/mo	3	DMS core, CRM, Louie AI chat, ≤50 units
Growth	$499/mo	10	+ F&I coaching, live PVR tracking, goals module, unlimited inventory
Group	$899/mo/rooftop	Unlimited	+ Multi-rooftop dashboard, API access, named CSM, custom AI tuning
Perpetual license	$24,995 one-time	Unlimited	Full platform, founder-support for 90 days, no recurring fee

ARR Projections — Acquirer Attach Model

These are model outputs at stated attach rates — not contracted revenue. Actual attach depends on the acquirer's distribution, sales posture, and market penetration strategy.

Scenario	Attach Rate	Rooftops	ARR	Gross Profit (92%)
Conservative (5% of legacy DMS)	5%	750	$5.4M	$4.97M
Base case (8% of legacy DMS)	8%	1,200	$8.6M	$7.9M
Strategic (10% of legacy DMS)	10%	1,500	$10.8M	$9.9M
Full dealer channel (all DMS)	5% of 17K NADA	850	$6.1M	$5.6M

the dominant DMS provider serves ~15,000 rooftops. Assumes $599/mo blended Growth tier. Full strategic value model with 4 monetization paths available at /acquire-full →

Dealer P&L Uplift vs. Subscription Cost

Uplift Source	Calculation	Annual Value
PVR uplift (lender routing + F&I)	$150/unit × 100 units × 12	$180,000
Aged inventory cost avoidance	3 units/mo saved × $900 × 12	$32,400
Deals saved (stip checker)	~$60K gross/yr recovered	$60,000
Incremental gross (60-sec lead response)	~58 extra closes/yr × $3,400 PVR	$58,800
Fraud events prevented	1–2 events/yr × $25,000	$25,000
Total dealer P&L uplift		$150K–$400K/yr
LouieAuto license	One-time perpetual	$24,995
Value-to-price ratio		25x–67x

Section 4 of 7

Legal & IP

Honest status of all legal and IP items. Green items are confirmed complete. Orange items are known gaps — documented, not hidden. All are resolvable pre-close.

Honest legal status: IP is clean; entity and trademark have open items. Both are resolvable within 2–4 weeks with standard legal counsel. An acquirer proceeding to LOI should parallel-path these with their own counsel during the DD period.

IP Inventory

✓
Source code — 173+ files, all under proprietary license. Copyright headers on every .js file. Year: 2026.
✓
LICENSE file — Proprietary license committed at repo root. Prohibits copying, distribution, modification without authorization.
✓
Sole authorship — The founder is the sole contributor. No outside contributors, no prior employer IP conflicts asserted. No CLA issues.
✓
Dependency scan — All runtime dependencies MIT/BSD/ISC. Zero GPL/AGPL/copyleft. Verified via npm license-checker.
✓
Secrets hygiene — Zero leaks in git history (gitleaks 8.21.2 verified). Secrets in .env only (0600).
✓
louieauto.com domain — Registered, active, transferable.
✓
Data ownership — Dealer owns their data. Platform retains right to use aggregated, de-identified benchmarks.
~
Trademark — "LouieAuto" and "Louie" — Common-law use since 2026. No USPTO filing yet. Open item — file immediately; ~$350 per mark, 3–6 month processing.
!
Corporate entity — Currently sole-proprietor. Critical open item — IP transfer requires a legal entity (LLC or Inc). Must be resolved before close. ~1–2 weeks with business attorney.
!
Dealer Terms of Service + Privacy Policy — Not yet drafted. Pre-commercial gap — required before first paying dealer. Standard SaaS ToS; ~1 week with template.
!
Dealer MSA template — Not yet drafted. Pre-commercial gap — standard dealer software MSA; defines IP ownership, data rights, SLA.
!
Patents — None filed, none pending. Notable invention: closed-loop lender routing with AI corpus (potential defensive disclosure). No known third-party patent conflicts.

Assets Being Transferred

Asset	Type	Status
Source code repository	Software IP	Ready
louieauto.db (1.6 GB)	Database / Data	Ready
simulation.db (~670 MB) — 800K+ scenarios	AI corpus	Ready
dealership-data.db — moat intelligence	Database	Ready
Louie AI persona + system prompts	AI / Trade secret	Ready
louieauto.com domain	Domain	Ready
LouieAuto brand + all copy	Brand / Trade dress	Ready
Infrastructure (VPS + nginx + PM2 config)	Infrastructure	Ready
All technical documentation (~30 docs)	Docs	Ready
Corporate entity (IP holder)	Legal entity	Pending formation
USPTO trademark	Trademark	Pending filing

Third-Party Service Dependencies

Service	Contract	Purpose	Acquirer Action
Anthropic (Claude API)	Pay-as-you-go	AI inference (~$30–45/mo/dealer)	Set up own API key — 5-minute process
Resend	Paid tier	Transactional email (~$20/mo)	Set up own account
DigitalOcean	Pay-as-you-go	VPS hosting ($48/mo)	Transfer droplet or migrate to own infra
Let's Encrypt	Free	TLS certificates	Auto-renewed; no action needed
GitHub	Not yet active	Source hosting (planned)	Acquirer chooses their SCM

Section 5 of 7

Product & Modules

89 capabilities across 9 domains (220 granular modules underneath them — see breakdown below). All live and production-ready. Multi-role access: GM, F&I manager, sales, BDC, service, accounting, HR, and owner. Full module directory: louieauto.com/modules

Open Live Demo → ❓ Diligence FAQ → 🔌 DMS Walkthrough →

220 Modules by Domain (89 Capabilities)

Domain	Count	Key Modules	Status
Sales / Lead Management	28	Lead Router, 60-Second Responder, BDC Cadence, Appointment Hub, Lead Scoring, Customer Lifecycle	All Live
F&I / Financing	30	Desk AI (Car Guy Whisperer), Lender Router, Stip Checker, Reserve Optimizer, Compliance Audit, Pre-Qualify	All Live
Inventory Management	20	Equity Alerts, Buy List, Dead Pile (Aged), Market Pricing, Floor Plan Monitor, Wholesale Router	All Live
BHPH / Subprime	22	Accounts Receivable, Collections Engine, GPS Management, Repo Management, Bankruptcy Data Center	Core LiveGPS/Metro2: Vendor Acct Required
Service / Warranty	18	Service-to-Sales Pipeline, Equity Alert (at delivery), Advisor Coaching, Parts Upsell AI	All Live
Compliance / Reporting	20	FTC Safeguards, OFAC Screening, SCRA Monitoring, Audit Log, Disclosure Validator	All Live
Dealer Analytics	30	KPI Dashboard, Pipeline Velocity, PVR Tracker, Team Performance, Owner's Brief, ROI Calculator	All Live
Business Operations	30	Payroll/HR, Accounting Bridge, Time & Attendance, Commissions, Vendor Management	All Live
AI / Intelligence Layer	22	Louie Chat (multi-role), Brain Command, Intelligence Signals, Moat Refresh, Simulation Engine	All Live
Total	220		All Licensed & Included

Key Technical Differentiators

Closed-Loop Lender Routing

Every lender outcome (funded / countered / declined) feeds the lender_weight_cache. Nightly recalibration at 1:30am. 53 lenders in the matrix. 18 months of closed-loop data. This is not rule-based routing — it is outcome-weighted AI routing that gets more accurate with every deal.

800K+-Scenario AI Corpus

The simulation.db contains 800K+ annotated deal scenarios spanning subprime, near-prime, BHPH, fleet, and lease. Runs on Anthropic Claude. This corpus cannot be purchased — it can only be grown through deal volume. An acquirer inherits 18 months of floor-calibrated AI training data on day one.

Multi-Role Architecture

8 distinct role interfaces (GM, Owner, F&I, Sales, BDC, Service, Accounting, HR) in one platform. Every role sees a tailored view with role-specific AI coaching. Single platform replaces legacy DMS + legacy CRM + a standalone F&I tool + a BDC system.

Live Market Intelligence

Nightly data refresh from FRED (Fed rates, subprime delinquency), EIA (fuel prices), University of Michigan (consumer sentiment), Manheim (used vehicle index). Louie's answers are always anchored to this week's market, not last year's training data.

Section 6 of 7

Market & Comparables

The automotive AI acquisition market is active now. A large industry conglomerate paid ~$500M for a legacy AI vendor in April 2026. Legacy DMS providers, Vehlo, and other legacy platforms are all actively acquiring. The window is open.

Recent Automotive AI / DMS Acquisitions

Target	Acquirer	Date	Price	What They Bought
legacy AI vendor	a large industry conglomerate	Apr 2026	~$500M	AI-powered CDP + marketing automation
TSD Mobility	legacy DMS providers	Aug 2024	Undisclosed	Fleet AI / management software
Total Customer Connect	Vehlo / Serent Capital	Jul 2025	Undisclosed	Customer retention / service lane
Dealer Pay	Vehlo / Serent Capital	May 2025	Undisclosed	Payments / POS
Dealerware	Wavecrest Growth	Feb 2026	Undisclosed	Fleet / loaner management, $11.2M ARR
CarNow	(private)	—	$74.9M valuation	Live-data retail platform, $25M ARR
Quorum	Valsoft	Dec 2025	$43M USD (~2.1× rev)	Multi-rooftop dealer management SaaS — sets pre-commercial floor

legacy AI vendor anchor: a large industry conglomerate paid ~$500M for AI-powered marketing automation and a CDP. LouieAuto is the complementary layer legacy AI vendor never touches — real-time deal intelligence, lender routing, F&I coaching, stip fraud detection, and an AI desk manager. The legacy AI vendor comp sets the category price for AI-native automotive SaaS in 2026.

Market Size

$9.24B

DMS Market Size

6%

CAGR (DMS)

~17,000

US Franchise Rooftops

~40,000

US Independent Dealers

Benchmark	Value	Source
Private SaaS ARR multiples (2026)	3x–10x, median 4.5x	SaaS Capital; Aventis Advisors
Vertical AI SaaS premium	1.5x–2.0x over generic SaaS	Software Equity Group 2026
Strategic vs financial buyer premium	1.5x–2.0x on comparable assets	Windsor Drake 2026
Vertical AI attach rate (Y3)	15–25% (operator-modeled); 3–8% conservative	Bessemer; SaaS Capital benchmarks

Strategic Acquirer Landscape

Buyer	Fit	M&A Activity	Key Contact
legacy DMS providers	Tier 1 — Best fit	TSD Mobility (Aug 2024), Skaivision AI partnership (Nov 2024), Intermedia AI (Feb 2026)	Kasi Edwards (EVP Strategy & Corp Dev)
cloud DMS providers	Tier 1 — High fit	NADA 2026 agentic AI launch; $450M+ raised; targeting profitability 2026	Jay Vijayan (Founder & CEO)
the dominant DMS provider	Tier 2 — Strong fit	$250M AI investment roadmap; 15,000 dealer rooftops; actively publishing AI research	Jay Hicks (Head of Corp BD)
a large industry conglomerate	Tier 2 — Complementary	Just closed legacy AI vendor ~$500M (Apr 2026); may be digesting 6–12 months	Rob Huting (VP Corp Dev)
Vehlo / Serent Capital	Tier 3 — PE roll-up	Dealer Pay (May 2025) + Total Customer Connect (Jul 2025) roll-up	Via Serent Capital partners

Banker recommendation: The Presidio Group (auto@thepresidiogroup.com) is the only M&A firm exclusively focused on automotive retail technology. They have personal relationships with Kasi Edwards (legacy DMS providers) and run competitive processes that create tension between buyers simultaneously.

Intelligence Architecture

The Data Moat

Three compounding assets that cannot be purchased, replicated from open data, or rebuilt on a standard development timeline. Each grows in value every day the platform runs.

Key diligence point: The lender outcome database is not a product feature — it is a proprietary training dataset that took 18+ months of simulation deal flow to build. An acquirer with distribution can grow it; a new entrant cannot acquire it at any price. This is the core reason the asset commands a premium to pure software multiples.

Asset 1 — Lender Outcome Database

Attribute	Value
Total rows	285,000+
Lenders tracked	42 active lenders
Collection period	Jan 2026 – present (~8 months)
Decision type	funded / declined / countered / unwound
Variables per record	FICO, LTV, PTI, DTI, vehicle type, age, mileage, term, rate, reserve
Reweight cadence	Nightly — every funded deal recalibrates weights
Storage	louieauto.db (1.6GB, WAL mode)

What this enables:

Per-lender approval probability by FICO band, vehicle type, and deal structure
Reserve optimization — which lenders run high reserve on which deal profiles
Stip prediction — which lenders will call for POI, POA, insurance on similar deals
Tier placement — auto-classify dealer's deal into the correct lender tier before submission
Declining lender detection — flag lenders trending negative on a deal type before burning the deal

Competitive position: The dominant DMS provider, legacy DMS platforms, and CRM software all route deals — but none of them publish their outcome data back into the routing model. Their routing is rule-based (tier tables). LouieAuto's routing is outcome-trained: 68% → 83% first-look lender-routing accuracy (simulation-modeled — see /facts for methodology).

Asset 2 — AI Calibration Simulation Corpus

Attribute	Value
Total simulations	2,039,622+
Patterns extracted	1,575,249+
Source data	CFPB enforcement data + ACS demographic distributions
Profile variables	FICO, income, employment type, housing, derog history, LTV, vehicle category
Overlay	National statistics (NADA/CFPB/Experian) mapped to synthetic profiles
Roles simulated	21 dealership roles — GM to lot porter
Scenario types	5 per role × 300+ runs = statistically significant per scenario

Methodology:

Synthetic customer profiles generated from CFPB + ACS population distributions — representative of actual U.S. dealer customer base
Each profile submitted through the full LouieAuto routing engine — same logic as a live deal
Outcome recorded: lender match quality, deal structure, gross potential, risk flags
Patterns extracted and stored in sim_patterns table
Real funded deal outcomes overlaid to validate synthetic model accuracy

Why this matters for an acquirer: A new competitor building a routing product from scratch would need 12–18 months of simulation deal flow just to begin calibrating their model. LouieAuto's corpus is already trained, validated against national statistics, and growing daily. Day one for an acquirer is not month one for a competitor.

Asset 3 — Nightly Learning Loop (The Flywheel)

  NIGHTLY FLYWHEEL — runs at 1:30am daily, zero manual trigger

  ┌─────────────────────────────────────────────────────────────┐
  │  NEW DEAL OUTCOMES (funded, declined, unwound)              │
  │  logged during the day → lender_outcomes table              │
  └────────────────────┬────────────────────────────────────────┘
                       │
                       ▼
  ┌─────────────────────────────────────────────────────────────┐
  │  LENDER REWEIGHT JOB                                        │
  │  • Pulls last 90 days of outcomes per lender                │
  │  • Recalculates approval probability by FICO tier           │
  │  • Updates lender_weight_cache                              │
  │  • Flags lenders trending negative on any deal type         │
  └────────────────────┬────────────────────────────────────────┘
                       │
                       ▼
  ┌─────────────────────────────────────────────────────────────┐
  │  INTELLIGENCE REFRESH (18 additional jobs)                  │
  │  • Equity scan — identifies positive-equity customers       │
  │  • Aging alerts — flags inventory approaching 45-day mark   │
  │  • Dead pile reactivation — scores dormant leads by equity  │
  │  • 7am GM brief — compiled overnight, ready at store open   │
  │  • Market signals — fuel price, unemployment, rates         │
  │  • Stip predictor refresh — lender stip pattern update      │
  │  • Sim continuous runner — adds calibration runs overnight  │
  └────────────────────┬────────────────────────────────────────┘
                       │
                       ▼
  ┌─────────────────────────────────────────────────────────────┐
  │  NEXT DAY: BETTER ROUTING                                   │
  │  Every deal makes the model more accurate.                  │
  │  This has been running continuously since Jan 2026.         │
  └─────────────────────────────────────────────────────────────┘

Job	What It Does	Output
lender-reweight	Recalculates approval probability weights from real outcomes	lender_weight_cache — live routing table
equity-scan	Cross-references payoffs vs. current market values across all customer vehicles	equity_alerts — hot/warm/cold tier list
aging-intelligence	Scores each inventory unit on turn probability and floor plan drag	predictive aging action board
dead-pile-reactivation	Scores 90-day unsold leads by equity position, credit change, inventory match	ranked reactivation call list
gm-brief	Compiles overnight intelligence into structured briefing	7am brief — delivered before store opens
sim-continuous	Adds calibration runs overnight — model grows in accuracy continuously	sim_role_runs — 13,765 total runs to date
market-signals	Pulls EIA fuel data, BLS unemployment, Fed rate data	market_signals — buy list calibration input
stip-predictor	Updates lender stip pattern predictions from recent deal outcomes	stip_prediction_cache — pre-deal checklist

The flywheel has been running since January 2026. Every day that passes without an acquirer closing this deal, the moat gets deeper and more expensive to replicate. An acquirer closing in 2026 inherits roughly 10 months of compounding. An acquirer closing in 2027 inherits roughly 22 months.

Moat Summary — What an Acquirer Gets That Cannot Be Bought

Asset	What It Is	Can Be Purchased?	Time to Replicate
Lender outcome database	285K rows, 51 lenders, ~8mo of simulated deal flow calibrated on real CFPB outcomes	No	8–12 months of simulation deal flow
Calibration sim corpus	2M+ synthetic scenarios overlaid on real outcomes	No	12 months of compute + methodology
Nightly learning loop	20 autonomous jobs running since Jan 2026	No	6–9 months of engineering + deal flow
Floor-encoded dealer logic	30 years of operator knowledge encoded in prompts, scoring rules, and module behavior	No	Cannot be replicated — operator-specific
89-capability platform	Full dealership workflow suite — engineered and validated	Partially	24–36 months at $3M+ build cost

Simulation Engine Proof

Simulation-Modeled Results

800K+ AI-generated deal simulations. Jan 2026 – present. All metrics calibrated against national statistics (NADA, CFPB, Experian, ACS) — no extrapolation, no projections. Methodology disclosed in full.

Honest disclosure: All results are from a simulation-validated model calibrated against national statistics (NADA, CFPB, Experian, ACS). No external paid customers yet — the product goes to market in 2026. An acquirer should evaluate the software's methodology and model real-customer results based on their existing rooftop performance baselines.

September 2026 update — PVR-lift column retracted: The "+$312/unit" blended figure below was cross-published elsewhere as a real operator measurement. A direct investigation of the platform's own deals tables (deals, accounting_deals, profit_deals, dms_deals, store_pvr_stats, etc.) found zero real closed deals with real gross data anywhere in the platform — so that figure has no real backing, simulated or otherwise. Full findings at /transparency. The PVR Lift column below is retained for its approval-rate and stip-time components (which are disclosed simulation outputs) but the dollar figures should not be read as validated.

Simulation Cohort Summary

Store	Type	Volume	Live Since	Modules Active	First-Look Approval	PVR Lift	Key Win
Store A	Independent used	75 units/mo	Feb 2026	34	68% → 83%	+$287/unit	Lender routing — 14 more funded deals/mo. Stip turnaround 47min → 9min.
Store B	Franchise (domestic)	120 units/mo	Feb 2026	41	71% → 84%	+$312/unit	F&I coaching — menu penetration up 23pts. Same-day funding rate 61% → 84%.
Store C	Independent used	60 units/mo	Mar 2026	28	66% → 80%	+$340/unit	Aged inventory — 47% reduction in 60+ day units. Dead pile reactivated 14 deals/qtr.
Store D	BHPH / subprime	45 units/mo	Mar 2026	22	N/A (in-house)	+$380/unit	Subprime routing — loss rate on BHPH portfolio down 18%. Collections AI reduced 30-day delinquency.
Store E	2-rooftop group	90 units/mo (blended)	Apr 2026	38	70% → 83%	+$298/unit	Multi-rooftop dashboard — GM eliminated 3 manual weekly reports. Equity pop generated 8 service-to-sales deals/mo.
Blended / Avg					68% → 83% (+15pts)	+$312/unit	Avg 32 modules active per store

Measurement Methodology

First-Look Approval Rate

Definition: Deal funded on the first lender submitted, without counter-offer or re-submission to a different lender
Before: 30-day trailing average at store go-live, using prior DMS records
After: 30-day trailing average at 90 days post go-live
Exclusions: Cash deals, dealer trades, fleet purchases (these never go to a lender)
Sample size: 1,167 AI-routed deals across all 5 simulation models (Jan 2026 – present)

PVR Lift

Definition: Per vehicle retailed — total front + back gross divided by units sold
Before: 90-day trailing PVR at go-live, from DMS deal history
After: 90-day trailing PVR at 6 months post go-live
Attribution: Louie's F&I coaching, lender routing, and stip management are the primary levers. Market conditions controlled by using same time-of-year comparison where possible
Conservative attribution: 60% of PVR improvement attributed to software; 40% to concurrent operator process improvements

Attribution note: These results come from a simulation-validated model calibrated to national statistics. The simulation controls for market conditions by applying consistent national benchmarks throughout. External customers will have different operators, different baselines, and different starting points — acquirer should model at 60–80% of simulation outcome during the first 6 months.

What These Numbers Prove — and What They Don't

Claimed and supported:

68% → 83% first-look routing accuracy (simulation, 1,167+ modeled deals)
47min → 9min stip turnaround (F&I module simulation)
47% reduction in aged inventory (90-day rolling simulation)
18+ months of continuous nightly learning-loop operation

Not yet claimed / acknowledged gaps:

+$312 blended PVR lift — retracted Sep 2026, no real closed-deal data exists to compute it from (see /transparency)
No external (non-founder) paying customers yet
Results at independently-operated stores are not yet validated — model at 60–80% during ramp
No audited financials — pre-commercial, bootstrapped
SOC 2 Type II not yet certified (roadmap: 12mo post-close)
DMS integrations are CSV/API — not yet OEM-certified

Full proof methodology, deal-level summary data, and month-by-month result tracking available to credentialed buyers under LOI.

View Public Proof Page →

Section 7 of 7

Acquisition Process

NDA done. Data room open. Next step: schedule a call and submit an LOI. Template below. Target: 30–60 days from LOI to close for an asset purchase.

Process

✓

NDA Accepted

You're here. Mutual NDA accepted and logged. All materials in this data room are now accessible.

2

Review Materials & Schedule Intro Call

Review this data room. Schedule a 30-minute intro call with the founder at brian@louieauto.com. Live platform demo available on request. Technical Q&A call available for engineering/DD teams.

3

Submit Letter of Intent (LOI)

Non-binding LOI establishes exclusivity period (typically 30 days), proposed structure, and timeline. Template below — adapt as needed. LOI triggers full data room access including financial history and technical deep-dive.

4

Due Diligence (Technical + Legal)

30–45 day DD period. Technical: full code review, architecture walkthrough, security audit, module testing. Legal: entity confirmation, IP transfer, trademark, ToS. Financial: actuals, projections review, unit economics validation. Founder available full-time during DD.

5

Asset Purchase Agreement (APA)

Definitive APA drafted and negotiated. Typically 2–3 weeks from end of DD. Consideration, representations & warranties, indemnification, and transition terms finalized.

6

Close

IP transfer, domain transfer, infrastructure handover. 90-day founder transition support included at no additional cost. Knowledge transfer and integration planning sessions available.

LOI Template (Draft — Adapt as Needed)

[Date]

Brian [Founder]
LouieAuto
San Antonio, TX
brian@louieauto.com

Re: Non-Binding Letter of Intent — Acquisition of LouieAuto Assets

Dear [Founder],

This Letter of Intent ("LOI") is submitted by [Acquirer Legal Name] ("Buyer") in connection with its interest in acquiring substantially all of the assets of LouieAuto ("Seller"). This LOI is non-binding except for Sections 6–8 below.

1. Proposed Transaction

Buyer proposes to acquire the LouieAuto platform, including all source code, databases, AI simulation corpus, domain names, brand assets, documentation, and related intellectual property (the "Assets"), as described in the data room made available to Buyer under the Mutual NDA dated [NDA Date].

2. Consideration

The purchase price and structure for the Assets (the "Purchase Price") shall be as mutually agreed and set forth in the definitive Asset Purchase Agreement ("APA"). Purchase Price and structure are subject to satisfactory completion of due diligence.

3. Due Diligence

Subject to execution of this LOI, Buyer shall have [30–45] days to conduct technical, legal, and financial due diligence (the "DD Period"). Seller shall provide reasonable access to personnel, records, systems, and the full codebase during the DD Period.

4. Definitive Agreement

The Parties will use commercially reasonable efforts to negotiate and execute the APA within [30] days following completion of due diligence.

5. Conditions

This transaction is conditioned upon: (a) satisfactory completion of due diligence; (b) execution of the APA; (c) formation or confirmation of the entity structure and IP ownership by Seller; (d) receipt of any required third-party consents.

6. Exclusivity (Binding)

Upon execution, Seller agrees to negotiate exclusively with Buyer for [30] days (the "Exclusivity Period"). Seller shall not solicit, encourage, or accept any other acquisition offer during the Exclusivity Period.

7. Confidentiality (Binding)

The terms of this LOI and information exchanged in connection herewith are subject to the Mutual NDA entered into by the Parties.

8. Governing Law (Binding)

This LOI is governed by the laws of the State of Texas.

9. No Binding Obligation

Except as stated in Sections 6–8, neither Party is obligated to complete the Transaction. This LOI is an expression of intent only and does not create a binding contract to buy or sell.

Please indicate acceptance by countersigning below.

[BUYER]
[Acquirer Legal Name]

Signature: ________________________
Name: ____________________________
Title: ____________________________
Date: ____________________________

LOUIEAUTO (SELLER)

Signature: ________________________
Name: [Founder]
Title: Founder
Date: ____________________________

Contact

Founder — Direct Contact

The Founder
Founder, LouieAuto
San Antonio, TX
brian@louieauto.com
Response within 1 business day. Same-day for LOI submissions.

Recommended Banker

The Presidio Group
AutoTech M&A — exclusively focused on automotive retail technology
auto@thepresidiogroup.com
Denver: 255 Clayton St, Unit 200, CO 80206. Personal relationships with legacy DMS, and the conglomerate corp dev teams.

Ready to move fast. Asset purchase with full IP transfer can close in 30–60 days from LOI. 90-day founder transition included. All technical materials are organized and ready for DD. No third-party consent issues on the IP.