Fortifai Limited · ASX:FTI
Fortifai — Investor Presentation (June 2026)
16 slides from Fortifai's investor presentation, Jun 23, 2026
Slide 1: Powering Agentic AI
The AI data plane CEO Presentation June 2026 Powering Agentic AI ASX: FTI
Slide 2: The team driving the Agentic AI infrastructure shift
INFRASTRUCTURE INNOVATORS The team driving the Agentic AI infrastructure shift Alon Rashelbach Kelly Herrell Jamie Pope Co-Founder & CTO CEO Chief Solutions Officer PhD in Computer Systems and Electrical Engineering. 3x Silicon Valley CEO. Three decades pioneering and Built high-performing PreSales
Slide 3: Agentic AI: Demand Explosion Begins
NEW MARKET Agentic AI: Demand Explosion Begins ‘Agents are the biggest unlock we’ve seen.’ 2022 2023 — 2024 2025 + ChatGPT changed the interface Investment cycle begins Agentic Big Bang ‘We are entering a phase of ‘always-on’ background agents that perform tasks when they are needed. It’s like takin
Slide 4: A Multi-Faceted Problem: Getting Data to the Agents
THE PROBLEM A Multi-Faceted Problem: Getting Data to the Agents Facet 1: The Data Path Facet 2: GPU Starvation Agentic AI flips the workload onto CPUs – the orchestration of multi-step, multi-function workflows, tool calling & governance. The GPU you paid for idles near 50%. Enterprise Agentic AI Us
Slide 5: Agentic AI at Scale Demands a Fundamentally New Data Plane
THE OPPORTUNITY Agentic AI at Scale Demands a Fundamentally New Data Plane Agent to Agent Communication and Governance Structured & Pre-Embedding + Inference Post-Inference Unstructured Data Pre-Inference Agent Outputs & Agent to Agent Traffic Loop Back – continuously, at wire speed Three Forces, al
Slide 6: Nol8: The AI Data Plane Efficient and Fast at Scale
EFFICIENCY SCALED Nol8: The AI Data Plane Efficient and Fast at Scale Nol8’s April 2026 Benchmarking Proved at All Workloads, Nol8 Can Realize Exponential CPU Savings ~500 servers (13 racks) 3 – 6 GWh / year 1 Accelerated Server 60,000 CPUs * Server power + cooling A t A ll Co mp lexities Legacy sys
Slide 7: Providing Economic Outcomes for the Agentic Economy
T H E AI D AT A P L AN E Providing Economic Outcomes for the Agentic Economy NeoCloud Economics Enterprise Economics GPU Infrastructure Owners Deploying Agentic AI at Scale Without Nol8 FRO M TO ≈50% Idle GPU Cycles $$$ $ cost per workload Affordable Scale Out With Nol8 High Latency Wire Speed GPUs
Slide 8: The Agentic AI Stack
T H E AI D AT A P L AN E The Agentic AI Stack Agentic Workloads WORKLOADS IN, RESULTS OUT Storage Compute DATA IN DATA OUT AI Data Plane Filter · Classify · Govern · Route Wire-Speed Throughput & Governance Network ASX: FTI
Slide 9: A $120B TAM by 2030
MARKET SIZE A $120B TAM by 2030 A · V AL U E P OO L ~$1.1T AI-infra spend × ~11% data-movement layer ≈ $120B TWO M ETHODS CONVERGE B · A CC O UNT - B AS E D $120B ~25,000 orgs with production AI × ~$4.5M ACV ≈ $113B 20 26 – 20 30 A GE NT I C T O K EN GR O W T H : 9 6% CA GR 1 IDC Worldwide Quarterly
Slide 10: We scale through NeoCloud integration…
GO-TO-MARKET We scale through NeoCloud integration… NeoCloud revenue ~$20B today, heading to ~$180B by 2030 = ~73% CAGR NeoClouds Nol8 AI Data Plane ASX: FTI https://www.srgresearch.com/articles/neoclouds-currently-growing-by-over-200-per-year-will-reach-180-billion-in-revenues-by-2030
Slide 11: …Then Distribute to their Customers as a Service
GO-TO-MARKET …Then Distribute to their Customers as a Service Neocloud Neocloud Neocloud Neocloud Neocloud Cost-effectively reach the AI-native builders Scaling Inference and Agents, globally. ASX: FTI
Slide 12: Pricing Model: Deliver Efficiency, Absorb the Profit Pool
GO-TO-MARKET Pricing Model: Deliver Efficiency, Absorb the Profit Pool Cost P RI C I N G $ per Gigabit-Second CPU-Based Consumption-Based: Customers pay for Throughput as they Scale. High TCO No ROI RE VEN U E Monthly Recurring Revenue Predictable, compounding MRR — not one-off hardware. MA RG I NS
Slide 13: A Wide and Deep Moat
COMPETITIVE DEFENSIBILITY A Wide and Deep Moat Three Layers, Co-Designed and Integrated into One AI Data Plane WH Y I T’ S DE FEN SI BLE Neural Network Algorithm The moat is the Deterministic transformations & policy enforcement — at line rate full stack — not any one layer. End-to-end integration S
Slide 14: AI Data Plane: The Missing Layer for Agentic AI
MISSING LAYER AI Data Plane: The Missing Layer for Agentic AI Full GPU Utilization ✓ Monetize the most expensive component in the datacenter Enterprise Agentic AI Users Storage GPUs Agent-Specific Data Delivery Data at Rest Data in use ✓ Feed curated, governed data for millions of agents Network Low
Slide 15: The AI data plane isn't optional. It's inevitable.
The AI data plane isn't optional. It's inevitable. ASX: FTI
Slide 16: Disclaimer
Disclaimer Authorised for release by the Board of Directors of FortifAI Limited. NOT AN OFFER OF SECURITIES Statements contained in this document, including but not limited to those regarding the possible or assumed future costs, performance, dividends, returns, revenue, exchange rates, potential gr