🧠
Investor Deck — April 2026

Luca AI Express

The AI-Native Enterprise Platform Nobody Else Has Built

An AI-native enterprise platform built on two decades of infrastructure experience across NVIDIA, Dell Technologies, and Microsoft, and a decade of focused AI engineering.
176+ production apps. NVIDIA Inception accepted.

20 yrs
Enterprise Experience
10 yrs
AI Engineering
176+
Production Apps
22
Business Domains
NVIDIA
Inception Member
K8s
Cloud-Native

Confidential — For Qualified Investors Only

The Opportunity

A Market Gap Nobody Has Filled

Enterprise software splits into two camps that don't overlap. Luca is the only product at the intersection.

Camp 1 — Legacy + Bolted AI

Salesforce ($330B), HubSpot, Zoho, ServiceNow. Deep features, massive ecosystems — but AI is a layer on top, not the foundation. Years of technical debt prevent true AI-native architecture.

Camp 2 — AI-Native, No Business

Dust.tt, Relevance AI, Dify.ai build excellent AI agent frameworks — but they're tools, not platforms. No CRM, no comms stack, no ERP, no app ecosystem. Can't replace a business suite.

Luca AI Express

Simultaneously AI-native AND a full business suite AND self-hostable. This isn't a marketing claim — it's a verifiable architectural fact. No other product occupies this intersection.

Capability Salesforce Odoo Dust / Relevance AI Luca AI Express
AI-Native ArchitectureBolted onMinimalYes✓ Yes
Full Business Suite (CRM, Comms, ERP)YesYesNo✓ Yes
Self-Hosted / On-PremNoYesNo✓ Yes
Multi-Channel AI CommsPartialPartialNo✓ Yes
Autonomous AI PersonasAgentforceNoYes✓ Yes
Kubernetes-NativeNoNoNo✓ Yes
What's Built

A Complete AI Operating System — Live & Deployed

Not mockups. Not wireframes. 176+ production apps running on Kubernetes right now.

49
OS Kernel Services

Boot, auth, RBAC, DB facade, AI engine, policies, ConfigBridge, telemetry

💬
60
Chat & AI Modules

Streaming, tool-use, RAG, skills, guardrails, code sandbox, integrations

🌐
22
Business Domains

CRM, comms, docs, finance, HR, analytics, security, media, training

🎨
36
UI Components

Dual-mode shell (desktop + mobile), web components, zero framework lock-in

🎤
5
Voice Pipeline Modules

WebRTC, STT, TTS, IVR routing, push-to-talk in 5 languages

🧠
8
Cognitive Persona Modules

Brain engine, attention system, comms layer, worker pools

🤖
7
Agent Runtime Modules

Claude SDK, terminal sessions, jobs, MCP protocol, autonomous builder

📩
7 ch
Communication Channels

WhatsApp, Telegram, Signal, Discord, WebChat, SMS, Email

Core AI

AI That Thinks, Not Just Responds

Every AI interaction runs through a cognitive architecture — not a simple prompt-response loop.

🧠 Cognitive Brain Engine (P8)

PERCEIVE → REASON → PLAN → ACT → REFLECT

Each persona runs a 5-phase cognitive cycle per interaction. Every phase makes a dedicated LLM call, creating deep understanding rather than shallow pattern matching.

💡 Perception Phase

Analyzes input for key facts, entities, intent, and emotional tone. Extracts structured context from unstructured messages.

🔬 Reasoning Phase

Draws conclusions from perception. Identifies knowns vs. unknowns. Cross-references with persona memory and conversation history.

📋 Planning Phase

Proposes 1-3 concrete action steps. Considers tool availability, user permissions, and domain constraints before acting.

⚡ Action Phase

Executes the plan — calls tools, queries databases, sends messages, triggers workflows. Produces the final response.

🙔 Reflection Phase

Reviews the entire cycle, extracts learnings, updates persona memory. Enables continuous self-improvement over time.

🔗 Supporting AI Infrastructure

🧰 Multi-Provider LLM Bridge (P3)

8 AI providers, automatic fallback chains, GPU queue integration (NATS), per-app config overrides. Model-agnostic — rides every AI wave (GPT-5, Claude, Llama, Gemini) without re-architecture.

🤖 Agent Runtime (G1)

Unified execution engine for all AI agents. Claude SDK integration, structured loop (Context → Plan → Act → Verify → Report), token budgeting, RBAC enforcement, session persistence with crash recovery.

💚 Healing Agent (P10)

Autonomous bug resolution pipeline. Detects errors in real-time, diagnoses root causes, proposes fixes, and can auto-repair issues — AI maintaining AI.

🎤 Voice AI Pipeline

Real-time voice with WebRTC + OpenAI Realtime. AI personas answer actual phone calls, handle IVR routing, support push-to-talk, in 5+ languages. Sub-second response latency.

💫 Dream Cycles

Optional periodic reflection where personas review recent interactions during downtime — extracting patterns and improving future responses. Continuous learning without explicit training.

📩 7-Channel AI Comms (OpenClaw)

WhatsApp, Telegram, Signal, Discord, WebChat, SMS, Email. Unified inbox with per-persona routing. AI handles conversations across all channels simultaneously.

Infrastructure

Kubernetes-Native Architecture: 100 to 1M Users

Not running on K8s. IS K8s. Custom Resource Definitions make the platform self-scaling.

Layer 3 — AIOS Custom Resources (CRDs)
AIOSDomain AIOSPersona AIOSAgent AIOSTenant

Domains, personas, and agents are K8s-native objects. kubectl to manage AI.

Layer 2 — AIOS Operator (Reconciles CRDs)
Domain Controller Persona Controller Scale Controller

Watch resources → auto-create schemas, mount routes, allocate memory, scale pods.

Layer 1 — Kubernetes Primitives
HPA / KEDA CloudNativePG Karpenter Helm NetworkPol
Layer 0 — AKS Infrastructure
Gateway ×3 GPU (NVIDIA A2) Comms ×2 Terminal Database Brains ×2

Key Infrastructure Capabilities

🛠 Tiered Boot System

Tier 0: Bare kernel (Express, DB, auth) — platform starts.
Tier 1: OS services (admin, secrets, policies) — manageable.
Tier 2: AI runtime (brain, personas, healing) — AI runs.
Tier 3: Business domains (CRM, chat, etc.) — full platform.
Dry deploy possible at any tier.

🗃 PostgreSQL 17.4 — Enterprise Grade

CloudNativePG operator, 3 replicas, 500Gi storage. Schema-per-domain isolation (crm.leads, comms.emails). PgBouncer connection pooling in transaction mode. Fully managed backups.

☁️ 6 Specialized Node Pools

Gateway (3 nodes), GPU inference (NVIDIA A2), Comms (2), Terminal, Database, Brains (2). Workload isolation by purpose — AI doesn't compete with database for resources.

🔒 Security-First Design

RBAC: {domain}:{resource}:{action} permissions. Auth on every route. Parameterized SQL only. XSS protection via esc(). Network policies. Secrets management via KeyVault.

🌐 Cloudflare Tunnel

Zero-trust ingress. No open ports. www.lucaexpress.com routes through Cloudflare tunnel to K8s ingress. Global CDN, DDoS protection, SSL termination included.

Architecture

172 Composable LEGO Pieces

Every module has a contract, declared dependencies, events, health checks. Plug in, plug out, scale independently.

OS Module Categories (49 Modules)

Category#Purpose
K — Kernel6Boot, registry, loader, event bus, config, scheduler
S — Shared Services9DB, auth, audit, cache, storage, sessions, notifications, events, secrets
R — Runtime8App registry, service manager, route proxy, health monitor, menu builder
I — Infrastructure7PostgreSQL, Azure, Cloudflare, Twilio, SMTP, Stripe, OpenClaw
L — Libraries6Text parser, markdown, rate limiters, connection tracker, helpers
P — Policies & AI10Token engine, guardrails, brain engine, persona runtime, healing agent
G — Agent7Agent runtime, terminal, job engine, MCP, builder, policy engine

Module Contract (Every Module)

export default {
  id: 'brain-engine',
  category: 'policies',
  requires: ['database', 'bus', 'token-engine'],
  register(kernel) { /* attach services */ },
  bootstrap(kernel) { /* start timers */ },
  shutdown(kernel) { /* clean up */ },
  health() { /* real metrics */ },
}

Token Engine — 8 AI Providers

OpenAI (GPT-5.4) Anthropic (Claude) Google (Gemini) Meta (Llama) Mistral Azure OpenAI Together AI Local vLLM (GPU)

Automatic fallback chains, per-app overrides, GPU queue via NATS, cost tracking per request. Never locked to one provider.

Economics

Structural Capital Efficiency

A traditional team would spend $1.4M–$2.3M and 12 months building this. Luca's AI-native operating model produced it as already-shipped infrastructure — the same model continues to compound on every new feature.

Component Traditional Team Traditional Cost Luca Actual
OS Kernel (49 services)3-4 senior engineers, 6 months$225K-$300KBuilt ✓ In Production
Chat/AI Engine (60 modules)2-3 ML engineers + 2 backend, 6 months$300K-$375KBuilt ✓ In Production
Voice Pipeline (WebRTC + AI)2 specialists, 4 months$100K-$150KBuilt ✓ In Production
22 Business Domains5-8 developers, 6-12 months$375K-$900KBuilt ✓ In Production
UI Shell + 36 Components2-3 frontend devs, 4 months$100K-$150KBuilt ✓ In Production
Persona/Brain Engine2 AI engineers, 6 months$150K-$200KBuilt ✓ In Production
OpenClaw Comms Gateway2-3 integration engineers, 4 months$100K-$150KBuilt ✓ In Production
K8s Infra, Helm, CI/CD1-2 DevOps engineers, 3 months$50K-$100KBuilt ✓ In Production
Total Replacement Cost 15-25 engineers, 6-12 months $1.4M-$2.3M Already shipped

This capital efficiency is not a one-time advantage. It is the ongoing operating model — the same model that produced this platform compounds across every new feature, every customer-specific deployment, and every integration. Investors are funding the model, not the line count.

Defensibility

Six Layers of Competitive Moat

Advantages that compound over time, not erode.

🚀 AI-Native Head Start

Salesforce and Odoo need years to re-architect around AI. We started there. Every month they spend retrofitting, we spend deepening capabilities.

📈 AI-Built Economics

Development cost 20-40x lower than traditional teams. Faster iteration, lower burn, ability to undercut incumbents on price while maintaining margins.

🔒 Self-Hosted Advantage

Regulated industries (healthcare, finance, government, defense) need on-prem AI. Salesforce can't do this. We can. $50B+ addressable market segment.

🧠 Persona Network Effects

Every deployment generates training data, tool patterns, and workflow templates. The platform gets smarter with each customer — a defensible data flywheel.

🔗 Multi-Model Insurance

Not dependent on any single AI provider. Token engine routes across 8 providers with automatic fallback. We ride every AI wave without re-architecture.

📩 7-Channel Lock-in

Once personas handle calls, email, WhatsApp, Telegram, and Discord, switching costs are enormous. Every channel deepens retention.

The key insight: Competitors must choose between being AI-native OR being a business suite. Luca is both — and the architecture makes it impossible to bolt one onto the other.

IP & Industry Visibility

Patent Pipeline, Publications & Industry Voice

The $25K–$40K IP allocation in each tier funds a concrete roadmap of provisional patents, technical publications, and conference visibility — building a defensive IP moat and category leadership in parallel.

8 Provisional Patents In Preparation

  • • Cognitive Persona Orchestration (PERCEIVE-REASON-PLAN-ACT-REFLECT)
  • • Multi-Channel Autonomous Comms Gateway (per-persona routing)
  • • Schema-Per-Domain Multi-Tenant Isolation
  • • Provider-Agnostic LLM Bridge (runtime token routing)
  • • Real-Time Voice-to-Persona Bridge (WebRTC + Realtime)
  • • Single-Image Tiered AI Platform Architecture
  • • Autonomous Bug Resolution & Self-Healing Pipeline
  • • LEGO-Piece Algorithm Registry & Module Composition

5 Technical Publications Drafting / In Prep

  • White Paper: AI-Native Enterprise ArchitectureDraft v0.1 complete. Industry whitepaper on the 176-app, 22-domain platform.
  • Schema-Per-Domain Isolation in Production AI SystemsSubmission target: VLDB Industrial Track.
  • AI-Augmented Software Engineering: An Operating ModelSubmission target: IEEE Software, ICSE Industrial.
  • GPU Infrastructure Patterns for Self-Hosted AICo-authored with NVIDIA AI architects via Inception.
  • Cognitive Persona Design: Brain Engine PatternsTechnical paper on attention-weighted cognitive loops.

5 Conference Appearances Pipeline

  • AI Engineer Summit 2026"From CRM to Cognitive Workflow" · CFP open
  • KubeCon + CloudNativeCon"Self-Hosted AI Platform on K8s" · CFP open
  • SaaStr AnnualFounder talk on capital-efficient AI-native B2B
  • NVIDIA GTCJoint session through NVIDIA Inception
  • Press & Podcast TourTechCrunch, The Information, Latent Space, a16z AI
Why this matters to investors: Provisional patents establish priority dates and buy 12 months to file full applications — defensive IP without immediate full prosecution cost. Publications convert technical depth into category leadership and developer mindshare. Conference visibility compresses sales cycles by establishing the founder as a category authority. Together these create a moat that compounds independent of headcount.
Financials

Revenue Projections

$25K one-time license acquisition + $20/user/month recurring + $32K one-time onboarding (4-week minimum @ $200/hr) per new customer. Customer-deployed on their own infrastructure. Internal AI inferencing — no per-token cost — drives 85%+ license gross margin.

Metric Conservative Base Case Optimistic
Customers (Month 12)2560120
Avg. Seats per Customer100250600
License $ / Customer / Month$2,000$5,000$12,000
MRR (Month 12, license)$50K$300K$1.44M
ARR Run Rate (Y1 exit)$0.6M$3.6M$17.3M
License Acquisition Y1 ($25K × cust.)$0.625M$1.5M$3.0M
Onboarding Revenue Y1 ($32K × cust.)$0.8M$1.92M$3.84M
Total Y1 Booked Revenue$2.0M$7.0M$24.1M
Customers (Month 24)75200400
ARR Run Rate (Y2 exit, license)$1.8M$12M$57.6M
License Acquisition Y2 ($25K × net new)$1.25M$3.5M$7.0M
Onboarding Revenue Y2$1.6M$4.48M$8.96M
Total Y2 Booked Revenue$4.65M$20M$73.6M
Gross Margin (License)82%85%88%
Gross Margin (Services)88%90%92%
Implied Valuation (10× ARR Y2, license-only) $18M $120M $576M
$50B+
Self-Hosted AI TAM

Healthcare, finance, government, defense need on-prem AI platforms. Luca's customer-hosted model directly serves this segment.

85%+
License Gross Margin

Internal inference, no external token cost, no infra to operate. Industry-leading SaaS margin profile.

40%+
YoY Enterprise AI Growth

Enterprise AI spending accelerating. Early movers in customer-hosted AI capture disproportionate market share.

The Ask

Three Tiers, Scaled to What Each Unlocks

The platform is already built. Each tier funds a different stage of commercialization, with explicit milestones tied to capital deployed.

Founder Round

$250K
12.5% equity · SAFE @ $2M cap

9 months — founder + assistant + IP foundation

✓ Founder full-time ($75K equiv., 9 mo)
✓ 1 operations assistant ($36K, 9 mo)
✓ $25K marketing & outreach
✓ $25K IP, patents, publications
✓ NVIDIA Inception co-marketing
✓ Cloud + legal reserve ($89K)
✓ Target: 5–10 paying pilots

Seed (Recommended)

$500K
10% equity · Priced @ $5M post-money

9 months — Founder Round + India engineering team

✓ Everything in Founder Round
✓ 6-engineer India team ($36K each, 9 mo)
✓ Multi-region deployment (US + EU)
✓ 5 IP filings, 2 conference appearances
✓ Cloud + infra scaling ($80K)
✓ Expanded compliance reserve
✓ Target: 30 customers, $80K+ MRR

Seed+

$1M
10% equity · Priced @ $10M post-money

12 months — full team, sales engine, Series A position

✓ Founder + assistant, full year
✓ 6-engineer India team, full year ($216K)
✓ 2 worldwide sales reps + commission ($200K)
✓ $120K marketing engine
✓ $40K IP, patents, publications
✓ SOC 2 Type I + legal ($80K)
✓ Target: 80+ customers, $300K+ MRR
$250K

Founder Round

9 months. Founder full-time. 1 assistant. $25K marketing. $25K IP. NVIDIA Inception leverage. 5–10 paying pilots.

$500K

Seed

9 months. + 6-engineer India team. Multi-region (US + EU). 30 customers, $80K+ MRR. 5 IP filings.

$1M

Seed+

12 months. + 2 worldwide sales reps. $120K marketing. SOC 2 Type I. 80+ customers, $300K+ MRR.

Series A

Market Leadership

Year 2. 100+ customers. $1M+ MRR. International expansion (EU, LATAM, MENA). Institutional traction at $3M+ ARR run rate.

Investor Terms

Equity, Valuation & Standard Rights

Indicative terms. SAFE structure (Founder Round) follows the Y Combinator post-money template. Priced rounds (Seed, Seed+) use NVCA Model Documents.

Term Founder Round — $250K Seed — $500K Seed+ — $1M
Equity Stake12.5% on conversion10%10%
Valuation$2M post-money cap$5M post-money$10M post-money
Security TypeSAFE (YC post-money template)Preferred Stock (NVCA)Preferred Stock (NVCA)
Liquidation PreferenceStandard SAFE conversion1x non-participating1x non-participating
Board RightsInformation rights onlyBoard observer seat1 board seat (of 3)
Pro-Rata RightsOn conversion to SeedStandard pro-rataStandard pro-rata
Information RightsQuarterly KPI summaryMonthly KPI + quarterly financialsMonthly KPI + quarterly financials + annual audit
Anti-Dilutionn/a (SAFE)Weighted-average broad-basedWeighted-average broad-based
Option Pool (post-money)Created at Seed close10% post-money10% post-money
Founder Vesting4-yr vest with 1-yr cliff at Seed close4-yr vest, 1-yr cliff4-yr vest, 1-yr cliff
Closing logic: Founder Round closes first as a rolling SAFE. Seed and Seed+ are mutually exclusive priced rounds — the company will accept whichever closes first with a credible lead investor. SAFE holders convert at the lower of cap or discount on the next priced round. Terms shown are indicative and subject to final due diligence and negotiation.
The Builder

Gustavo Assuncao

Founder & Chief Architect

Two decades of enterprise infrastructure work, built across the companies that defined modern computing — NVIDIA, Dell Technologies, and Microsoft. The last ten years focused specifically on AI engineering: model deployment, inference optimization, GPU-accelerated systems, and the architectural patterns that make AI work at production scale.

AI-native enterprise software is not a feature you bolt on — it is an architecture you must commit to from the first line of code. Doing it well requires deep enterprise software experience to know what business workflows need, and deep AI engineering to know what the technology can credibly deliver. Few people have lived in both worlds for this long. Luca AI Express is the downstream consequence of that combined judgment.

NVIDIA Inception Program Member. Luca AI Express has been accepted into NVIDIA Inception — NVIDIA's selective accelerator program for AI-native startups. Admission requires technical due diligence by NVIDIA on architecture, AI roadmap, and engineering depth. Membership grants GPU compute credits and preferential H100/H200 inference pricing, direct access to NVIDIA AI architects for technical co-engineering, marketing visibility through NVIDIA's startup channels, and eligibility for NVIDIA Ventures consideration. For investors, this is third-party technical validation.

📞
+1 (786) 442-4789
gus@gusit.de
📅 Book a Meeting 💻 Live Demo 🔍 Review the Code 📈 Discuss the Opportunity

www.lucaexpress.com  |  Confidential — For Qualified Investors Only