The Thesis · The Infrastructure Shift
NOVO Decentralized Inference Infrastructure — the first building block of a global compute orchestrator
The Status Quo
A monopoly of hyperscalers and a single chip manufacturer controls the supply and price of AI compute.
The Promise
NOVO secures wholesale inference capacity in energy-efficient data centers and delivers it globally through an OpenAI-compatible API – standard inference at market price, sovereign EU inference at a fraction of the cost of the proprietary stack.
Phase 1 (today): wholesale inference arbitrage in energy-efficient regions. Phase 2 (from 2029): horizontal orchestration across GPU, training, and storage capacity.
The Problem · The Price Dictate
A Monopoly Dictates
the Price of Intelligence
The Structure
A monopoly of hyperscalers and a single chip manufacturer determines supply, access, and price for AI compute.
The Real Bottleneck
Ideas don't decide growth — power, capital, and Nvidia allocations do.
The Consequence
AI companies pay monopoly prices and accept waiting lists because no structural alternative exists.
The Solution · The Alternative
Infrastructure Without
Owning Data Centers
from $0.80
per 1M tokens (standard)
Consistent
streaming latency
0
artificial rate limits
Secured wholesale capacity in energy-efficient data centers — far below the cost base of classic hyperscalers.
The entry point is inference. The same sourcing and contracting logic carries the expansion into further compute categories — the structural core of the marketplace model: aggregating demand without owning the underlying asset.
The Product · The API
Integration Without
Migration Risk
The Integration
100% OpenAI-compatible REST API. Migration in a single line of code — no refactoring of existing applications.
The Architecture
Aggregated wholesale GPU capacity from data center partners — no owned hardware, no depreciation risk.
The Performance
No artificial throttling within booked capacity — contractually secured wholesale blocks are distributed dynamically across customer traffic, with transparently communicated burst allowances for load spikes.
The Market · Positioning
Three Categories of
Compute Providers
Hyperscalers
AWS, Azure, Google Cloud. Fully integrated stack, high prices, long waiting lists for current GPU generations, little price competition.
Inference Specialists
Together AI, Fireworks, Groq, DeepInfra. Efficient open-weight inference at competitive prices — without a structural cost advantage from energy location.
NOVO
Market comparison, Q2 2026 (Llama 3.3 70B class): Together AI $0.88/1M, Fireworks AI $0.90/1M, Groq $0.59 input / $0.79 output per 1M at industry-leading speed (250+ tokens/sec). NOVO stands at $0.80/1M tokens standard rate — on price parity with Together/Fireworks, without Groq's speed edge. The differentiator is therefore not raw price, but EU/GCC data residency, bundled enterprise SLAs, and dedicated capacity guarantees.
The Moat · Enterprise Security
Security as
a Design Principle
Zero-Retention by Design
Every request is processed in an isolated process environment (process isolation). Prompts are never stored, logged, or used for training.
Data Residency & SLAs
Optional EU data residency and dedicated routing secure contractual enterprise SLAs for regulated industries. By aggregating across multiple capacity providers, bundling SLAs, and offering sovereign routing, NOVO remains the more reliable layer than going direct to any single provider.
Risk & Resilience · Regulatory Framework
Capacity That Holds Up
Under the Export Regime
The Framework
The US export control framework for advanced AI chips into the Gulf region was expanded in November 2025 for select licensed entities (including HUMAIN in Saudi Arabia and G42 in the UAE), but remains license-bound and volume-capped.
Our Safeguard
Capacity sourcing exclusively through licensed, export-control-compliant regional partners, with contractually fixed end-use and compliance checks. Additional locations with comparable energy cost structures are continuously evaluated to diversify beyond the Gulf region.
The Structure · Capital & Capacity
The Bridge Between
Delaware and Riyadh
Capital access, intellectual property, and USD revenue base — prepared for a dual-track exit.
Operational hub for capacity sourcing, energy arbitrage, and B2B access across the GCC.
The Economics · Price & Margin
The Structural
Price Advantage
Market-rate pricing with a realistic 40% gross margin — EBITDA-positive from 2028, cumulative break-even in Q3 2029, asset-light with no data centers of our own. Model assumptions (customer count, utilization, price development) in the appendix.
Note: this comparison sets NOVO's open-weight inference cost against proprietary frontier model APIs — comparable output quality for standard enterprise workloads, not an identical model. Direct competitive comparison against open-weight providers (as of Q2 2026): Together AI $0.88/1M, Fireworks AI $0.90/1M, Groq $0.59 input / $0.79 output per 1M at superior speed. NOVO's $0.80/1M sits between Groq and Together/Fireworks on price — the competitive edge lies in data residency and SLA bundling, not a dominant price gap.
Scaling & Vision
A Network Effect
That Reinforces Itself
The Mechanism
More purchased capacity unlocks more demand — a self-reinforcing effect, secured by the LOI price freeze.
The Price Anchor
Every LOI customer locks in the price permanently — a structural incentive for early demand commitment.
2028
~$87M net revenue
2030
~$410M net revenue
~$2.3B
Modeled Series C valuation 2030
Assumptions (customer count, ARR/customer, multiple) in the appendix. Expansion stage: from Phase 2, the network effect also carries beyond inference into training and storage capacity — see roadmap.
The Roadmap · Milestones
4 Years to
Operational Maturity
Binding capacity allocation, API MVP, and first LOI customers.
SOC 2 and ISO 27001 certification, EU data residency, launch of enterprise beta.
Turn to positive EBITDA in live operations.
Series B preparation and evaluation of concrete exit paths.
Phase 1 (Year 1–2): the inference wedge. Phase 2 (Year 3–4): platform expansion. The Series C valuation shown on the growth slide sits beyond this 4-year horizon (Year 5+) and depends on successfully closing Series B in Year 4.
The Ask & Use of Funds
$35M Seed Round to
Secure Capacity and Market Access
Use of Funds
Ijara: a Sharia-compliant leasing instrument used to secure hardware-adjacent assets without additional equity dilution.
The NOVO Principle
The structural undersupply of inference capacity is real.
NOVO secures access to it — on terms not available in the market today.