Investor Presentation

E2E Networks Limited CIN- L72900DL2009PLC341980 Uppal's Genesis A-32, Block B, Mohan Cooperative Industrial Estate, Badarpur, New Delhi 110044, Phone No. +91-11-4084-4964 Email: cs@e2enetworks.com, Website https://www.e2enetworks.com/ Date: October 16, 2026 To, To, National Stock Exchange of India Ltd. (‘NSE’) The Manager- Listing Exchange Plaza, 5th Floor, BSE Limited, Plot No. C/1, G Block, Phiroze Jeejeebhoy Towers, Bandra - Kurla Complex, Bandra (E), Mumbai – Dalal Street, Mumbai- 400 001 400 051, Maharashtra, India Maharashtra, India (Script Symbol: E2E) (Scrip Code: 544783) Sub: Investor Presentation of E2E Networks Limited (“the Company”) Dear Sir/Ma’am, Pursuant to Regulation 30 of the Securities and Exchange Board of India (Listing Obligations and Disclosure Requirements) Regulations, 2015, please find enclosed herewith Investor Presentation of the Company. The aforesaid presentation is also accessible on the Company’s website at https://www.e2enetworks.com . This is for your information and records. Yours faithfully, For E2E Networks Limited Ronit Company Secretary & Compliance Officer Membership No.: A59215 Encl.: As above Q2 FY27 Investor Presentation Oct 16, 2026 Safe Harbour Statement This presentation has been prepared by E2E Networks Limited ("the Company") solely for information purposes and does not constitute an offer, invitation, solicitation or recommendation to purchase or subscribe for any securities, nor should it form the basis of, or be relied upon in connectionwith, any investment decision, contract or commitment. Certain statements contained in this presentation, including those relating to the Company's future business prospects, growth strategy, AI infrastructure capacity expansion, cloud platform capabilities, subsidiary operations, financial performance, capital allocation, and industry outlook, are forward-looking statements. These statements are based on management's current expectations, assumptions and estimates and are subject to known and unknown risks, uncertainties and other factors that may cause actual results, performance or achievements to differ materially from those expressed or implied by such statements. These factors include, but are not limited to, changes in economic and market conditions, demand for AI and cloud infrastructure services, competitive dynamics, technology advancements, availability and pricing of GPU infrastructure, execution of expansion plans, customer adoption, regulatory developments, cybersecurity risks, access to capital, and other risks beyond the Company's control. The Company undertakes no obligation to update or revise any forward-looking statements, whether as a result of new information, future events or otherwise, except as required under applicablelaw. Past performance should not be regarded as an indicator of future performance. Who We Are India’s Sovereign Provider of Advanced Cloud GPUs 16+ ~5,100 years of operational cloud GPUs live including 1,024 excellence NVIDIA B200s What’s New this Quarter @E2E | Q2 FY27 ₹1,000 Cr Sovereign Adding Leadership Private Networking Jarvislabs.ai AI Contract Heft Comes to TIR The AI Cloud Entered into a binding term Anand Ganapathy (ex-IBM, Added VPC Connect enables Managed Endpoints & sheet worth ~₹1,000 Cr (excl. Dell) joins as Chief Revenue secure, private networking for Model APIs: Production- Officer, bringing 27+ years of taxes) with an India-based GPU instances, training ready open-weight models enterprise sales leadership. Sovereign AI company for clusters and inference and pay-per-token APIs for NVIDIA Blackwell cloud GPUs Abhishek Shindadkar joins as endpoints, delivering open-source models. Head, Founder’s Office, and allied services valid up enterprise-grade isolation bringing 21+ years of expertise to June 2029. Validates long across the AI lifecycle for Infrastructure Upgrades: in technology, research and term visibility and ability to regulated workloads, at par InfiniBand H200 clusters, capital markets. transition deployed capacity with global hyperscalers. serverless inference that Vinod Kumar joins as SVP, Field from a pay-as-you-go scales to zero, private VPC – CTO, with 20+ years of model to a longer-term networking, Object Storage experience across cloud commitment. infrastructure, HPC, storage and shared filesystems. and enterprise technology. E2E at Global Fintech Fest 2026 E2E and Jarvislabs.ai exhibited at Global Fintech Fest 2026, Mumbai, presenting India-hosted GPU cloud and AI platforms to banks, insurers and fintechs. Engaged with fintech stakeholders on AI, cloud infrastructure AI & Cloud Scalable Industry and the growing need for scalable, reliable computing. Infrastructure Computing Connections A Structurally Undersupplied Market, Compounding Fast GPU demand for AI training and inference continues to outstrip hyperscale capacity — pulling capital into purpose-built neocloud providers worldwide Global Neocloud Market Revenue US$6.7T (US$ Billions) Global data center capital expenditure needed by 2030 —~70% driven by AI workloads $400B Source: McKinsey & Company, "The cost of compute" (Apr 2025) 100+ Neocloud / GPU-cloud providers active globally $180B today —a highly fragmented, fast-growing category Source: DriveNets; Thunder Compute (2026) $25B 2025 (Actual) 2030 (Forecast) 2031 (Forecast) share of neocloud market revenue expected to come from AI inference by 2030 Source: Synergy Research Group (2026) Source: ABI Research, "The State of Neocloud" (2026) Built in India, for India’s AI Future. Engineering Owned, Not Rented India based GPU cloud provider listed in NSE and BSE. Full stack, in-house — 100+ engineers building compute, networking, orchestration & AI software. Own AI platforms — two in-house platforms: TIR (training, fine-tuning, RAG & model endpoints) and Jarvislabs.ai (developer friendly GPU cloud). Sovereign compute — nearly 3,000 Hopper and 1,024 Blackwell GPUs deployed. Further, scaling Blackwell Capacity Multi-year Blackwell GPU contract worth ~₹1,000 Cr with an India-based sovereign AI company. INTEGRATED AI PLATFORM & MONETIZATION MODELS AI INFRASTRUCTURE Public Platforms – TIR & Jarvislabs.ai Training, fine-tuning & inference for LLM builders and NVIDIA GPUs (Hopper Blackwell) in Delhi AI-native teams at cluster scale NCR & Chennai DCs GO-TO-MARKET Sovereign / Private Platform Self-serve portal + enterprise sales + partners + E2E’s own cloud software deployed for regulated & government empanelment data-residency-bound customers TIR AI/ML Platform Sovereign AI, end to end — build, fine-tune, deploy & scale GPU Notebooks Foundation Studio Jupyter on H200 & B200 GPUs, NGC Fine-tune Llama, Mistral, Gemma & frameworks preloaded Stable Diffusion RAG & Vector DB Model Endpoints Managed Qdrant + embeddings, Auto-scaling inference via vLLM, chatbot live in minutes SGLang, Triton or PyTorch Pipelines & Training GenAI API Datasets, model repo, pipelines & LLM, speech & embedding models via multi-node training OpenAI-compatible API Jarvislabs.ai — Developer First GPU Cloud Spin up GPUs in seconds from CLI or code — an E2E Platform since December 2025 Instant GPU Managed Endpoints CLI, SDK & Pause & Resume Instances & Model APIs: Templates Launch H100, H200, A100 Pause an instance to pay and RTX Pro 6000 GPUs on only for storage —and Production-ready open- Manage the full instance demand, in seconds. switch GPU type when you weight models and pay- lifecycle from terminal or resume. per-token APIs for Open- Python, with framework Source Models. templates and persistent storage. Jarvislabs.ai: the developer-first GPU cloud — an E2E Platform since Dec 2025. The Growth Story Capacity meets Demand TODAY ON THE ROADMAP IN DEPLOYMENT Hopper & Blackwell Blackwell scale-up B300 / Vera Rubin GPUs ~5100GPUs live, (including 1024 B200s), 1024+ B200s planned Next-generation Compute Blackwell capacity now under multi- planning underway year contract GPU Trajectory 5,100 3,900 1,900 FY25 FY26 FY27Q2 [Showing first 8,000 characters — download PDF for full document]