Pinpole - Simulate AWS, GCP, or Azure architecture before deployment

Hey 👋 I'm a business architect by trade. For the better part of a decade I've sat in the room where cloud bills get explained to CFOs, where post-mortems get written for the board, and where engineering leaders try to reconcile a beautifully designed architecture with the $40K surprise it produced in month one. The pattern was always the same. A senior engineer would design something competent. It would pass review. It would deploy. And then something nobody had modelled - a promotional spike, a cold-start cascade, a DynamoDB WCU ceiling, a Cosmos DB RU misconfiguration - would turn a good design into an incident. The post-mortem would land on an engineer's desk. The line item would land on the CFO's. What struck me wasn't the failures. It was that the tools made the failures structural. Teams were designing in draw.io, costing in a vendor pricing calculator, deploying with Terraform, and load-testing with k6 after the infrastructure was live. Four tools, zero shared state, and every significant decision committed to real spend before anyone could validate it. The workflow itself was guaranteeing the bad outcome. From a business architecture lens, that's not an engineering problem - it's a design-time governance gap. Cost, performance, compliance, and resilience are all decided at design time and discovered post-deployment. The entire $14.5B FinOps market exists downstream of that gap, trying to recover waste that shouldn't have been created in the first place. PinPole was built to close it. One canvas across AWS, GCP, and Azure. Pre-deployment simulation from 10 RPS to 100M RPS across four traffic patterns. Live cost at every RPS. AI recommendations that surface the architectural flaw before the engineer has to live through it. Deploy via short-lived cross-account credentials, or export to Terraform. The approach evolved the way most good products do — narrower than we planned, then broader than we expected. We started AWS-only because that's where the deepest scars were. GCP engineers asked next, then Azure. What began as a simulation tool became a design-time control surface: the point in the workflow where cost, performance, and compliance decisions can still be changed cheaply. Free tier, no credit card, no cloud account required to start. The question I keep asking engineers and architects: when did you last commit real cloud spend to a design you hadn't simulated? If the honest answer is "last week," that's the problem we built this for. Would genuinely value feedback from anyone who's carried a post-mortem like the ones I've described - what would you want caught before it became your Monday morning? - Damian Hickey Co-founder, PinPole