Episodes

  • How Cloud Bills Now Include a Data Egress Tax on AI Agents

    How Cloud Bills Now Include a Data Egress Tax on AI Agents
    Jul 4 2026
    Episode 90 dives into a new line item appearing on enterprise cloud bills: data egress charges specifically triggered by AI agents. Lucas explains how agentic workflows—where AI models autonomously call APIs, query databases, and return results—multiply egress costs compared to traditional user-driven traffic. The hosts examine Amazon Bedrock's new 'Agent Data Transfer' fee, how it compares to Azure's API Management consumption tiers for AI, and why enterprises running multi-agent systems on Google Cloud are seeing bills spike 30-40% quarter-over-quarter. Luna shares a story from a fintech startup that saw its monthly egress bill jump from $4,000 to $18,000 after deploying a basic customer-support agent. The episode also covers mitigation strategies: caching responses, colocating agents with data, and negotiating flat-rate egress lanes. Fresh angle not covered in previous episodes (which focused on GPU fees, IPv4 exhaustion, fiber networks, etc.). #CloudComputing #AWS #Azure #GCP #AI #DataEgress #AgentFees #CloudBills #Technology #Business #FexingoBusiness #BusinessPodcast #CloudInfrastructure #AIWorkflows #CostOptimization #AmazonBedrock #AzureAI #GoogleCloud Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    7 mins
  • Why Cloud Bills Now Include a Reserved Capacity Deposit

    Why Cloud Bills Now Include a Reserved Capacity Deposit
    Jul 3 2026
    Lucas and Luna break down a new line item appearing on enterprise cloud invoices: the reserved capacity deposit. Unlike traditional reserved instances that offer a discount in exchange for upfront commitment, this deposit requires customers to pay a non-refundable fee to 'hold' compute capacity in a specific region, even before they provision any resources. The hosts trace the origin of this practice to the GPU shortage of 2023–2024, when AWS and Microsoft began charging deposits to queue for H100 clusters. What started as an AI-specific workaround is now spreading to general-purpose CPU instances in constrained regions like US-East and EU-West. Lucas explains the accounting implications — the deposit sits as a prepaid asset on the balance sheet, but if the customer never deploys, the cloud provider keeps the cash. Luna questions whether this is essentially a call option on compute. The episode closes with a practical tip: negotiate a sunset clause that lets you convert unspent deposits into credits before the fiscal year end. #ReservedCapacityDeposit #CloudPricing #AWS #MicrosoftAzure #GoogleCloud #GPU #H100 #EnterpriseCloud #CloudBills #InfrastructureCost #TechPodcast #Technology #CloudInfrastructure #ITSpending #FexingoBusiness #BusinessPodcast #CloudProviders #ComputeCapacity Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    9 mins
  • How Cloud Bills Are Adding a Network Reliability Fee

    How Cloud Bills Are Adding a Network Reliability Fee
    Jul 3 2026
    Episode 88 of Cloud Computing with Fexingo unpacks a new charge appearing on enterprise cloud invoices: the network reliability fee. Lucas and Luna explore why AWS, Azure, and GCP are introducing this cost tied to SLA guarantees and latency commitments. They use a concrete example of a mid-size e-commerce company whose monthly bill jumped 12% after a regional outage triggered the fee. The episode explains the engineering rationale—redundant fiber paths, diverse peering—and questions whether the fee is transparent or just another lock-in lever. Listeners learn how to audit their own contracts for this line item and negotiate credits. #CloudComputing #AWS #Azure #GCP #NetworkReliabilityFee #CloudBilling #SLAs #Latency #Egress #Infrastructure #Technology #FexingoBusiness #BusinessPodcast #CloudEconomics #VendorLockIn #EnterpriseIT #CloudContracts #NetworkEngineering Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    9 mins
  • How Cloud Bills Now Charge for Egress to Private Data Centers

    How Cloud Bills Now Charge for Egress to Private Data Centers
    Jul 2 2026
    In this episode, Lucas and Luna explore the latest cloud billing twist: egress fees for data moving from a public cloud to a customer's own private data center. They break down AWS's recent 2026 pricing change, which charges $0.09 per gigabyte for traffic exiting to on-premises infrastructure, even if the customer owns the network. The conversation covers why providers are pushing this, how it affects hybrid cloud strategies, and what companies like Netflix and Dropbox did to avoid similar costs. Lucas cites a report from CloudZero showing enterprises now spend an average of 18% of their cloud bill on egress, with private data center egress being the fastest-growing component. They discuss mitigation tactics like direct peering, caching, and re-architecting data flows, and question whether this accelerates the move to multi-cloud or pushes companies toward single-provider lock-in. The episode is practical, specific, and grounded in real numbers—anyone managing a hybrid infrastructure will learn something actionable. #CloudComputing #AWS #Azure #GCP #EgressFees #HybridCloud #CloudBilling #Infrastructure #DataTransfer #Networking #CloudEconomics #CloudZero #FexingoBusiness #BusinessPodcast #Technology #FinOps #MultiCloud #OnPremises Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    9 mins
  • How Cloud Bills Now Charge for Shared Accelerator Memory

    How Cloud Bills Now Charge for Shared Accelerator Memory
    Jul 2 2026
    Episode 86 of Cloud Computing with Fexingo dives into the latest line item on enterprise cloud invoices: shared accelerator memory. Lucas explains how AWS, Azure, and GCP are now charging for GPU and TPU memory that was previously bundled into compute costs. He breaks down the pricing model using NVIDIA H100 GPUs on AWS as a concrete example, showing how a single 80 GB H100 can now incur an extra $0.40 per GB per hour for memory reserved across instances. Luna questions whether this is a hidden price hike or a genuine reflection of supply constraints. The episode explores the infrastructure logic behind disaggregated memory, the impact on AI training budgets, and why this shift may accelerate adoption of memory pooling standards like CXL. A must-listen for any team managing cloud GPU workloads. #CloudComputing #AWS #Azure #GCP #GPU #TPU #NVIDIAH100 #SharedMemory #AcceleratorMemory #CXL #AIWorkloads #CloudBilling #Infrastructure #Technology #FexingoBusiness #BusinessPodcast #CloudEconomics #MemoryPooling Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    8 mins
  • How Cloud Bills Now Include a Regional Power Premium

    How Cloud Bills Now Include a Regional Power Premium
    Jul 1 2026
    Lucas and Luna dive into a new line item quietly appearing on cloud invoices: the regional power premium. As data center electricity costs vary wildly by location — from subsidized rates in Iowa to grid-strained markets in Northern Virginia — cloud providers like AWS, Azure, and Google Cloud have begun passing those differences through to customers. The episode examines how the premium works, why it's not a simple per-kilowatt-hour surcharge, and what it means for companies choosing where to deploy workloads. Lucas explains the three-tier pricing structure AWS introduced in early 2026, Luna pushes back on whether this is transparency or cost-shifting, and both discuss the implications for multi-region architecture decisions. A concrete look at how power prices are reshaping cloud strategy. #CloudComputing #CloudBilling #AWS #Azure #GoogleCloud #DataCenter #PowerPremium #ElectricityCosts #CloudStrategy #MultiRegion #Infrastructure #Technology #CloudEconomics #EnergyCosts #FexingoBusiness #BusinessPodcast #TechPodcast #CloudPricing Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    10 mins
  • How Cloud Provider Lock-In Is Getting Smarter and More Expensive

    How Cloud Provider Lock-In Is Getting Smarter and More Expensive
    Jul 1 2026
    Episode 84 of Cloud Computing with Fexingo digs into the evolving mechanics of cloud provider lock-in — specifically the 'stickiness' strategies that are harder to detect and costlier to escape. Lucas and Luna explore how hyperscalers like AWS, Microsoft Azure, and Google Cloud are embedding long-term commitments into managed services, data format dependencies, and AI platform integrations. They dissect a real example: the hidden costs of migrating away from a fully managed Kubernetes service when you've built custom operators and signed three-year commitment discounts. The episode also covers the rise of 'data gravity' charges and how newer fine-grained billing structures make it cheaper to stay than to leave. If you're running a multi-cloud or hybrid strategy in mid-2026, this episode gives you concrete numbers and negotiation angles to watch for. #CloudLockIn #AWS #Azure #GCP #ManagedServices #DataGravity #MultiCloud #HybridCloud #EgressFees #CommitmentDiscounts #CloudCosts #Kubernetes #AIInfrastructure #VendorLockIn #CloudBilling #FexingoBusiness #BusinessPodcast #Technology Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    8 mins
  • How Cloud Bills Are Adding a Storage Access Tier Surcharge

    How Cloud Bills Are Adding a Storage Access Tier Surcharge
    Jun 30 2026
    Episode 83 of Cloud Computing with Fexingo digs into a quietly proliferating charge on enterprise cloud invoices: the storage access tier surcharge. Lucas and Luna explore how AWS, Azure, and Google Cloud are rolling out per-request fees for accessing cold and archive storage, with rates that can multiply costs for frequent reads. The episode uses a concrete example: a healthcare company that stored five petabytes of medical imaging in AWS S3 Glacier Deep Archive, expecting to pay $3 per terabyte per month, but found its bill tripling after the introduction of the retrieval request fee for bulk operations. Lucas explains the engineering logic behind the pricing shift—cloud providers want to penalize hot access to cold data—and Luna questions whether the opacity of these tiers undermines the value proposition of cloud storage. They also discuss mitigation strategies like lifecycle policies and third-party caching layers. A single concrete takeaway: before moving data to a cold storage class, model your expected access patterns against the new per-request fees, not just the per-gigabyte storage rate. #CloudStorage #StorageAccessTier #AWSS3 #AzureBlob #GoogleCloudStorage #CloudBilling #ColdStorage #ArchiveStorage #RetrievalFees #InfrastructureCost #CloudEconomics #DataStorage #Technology #CloudComputing #FexingoBusiness #BusinessPodcast #CloudEngineering #FinOps Keep every episode free: buymeacoffee.com/fexingo
    Show More Show Less
    12 mins