We built Finopscraft because cloud bills became unreadable.
In early 2025, Megan was running engineering at a Dallas-based SaaS company where AWS spend had tripled in six months. Nobody could say which service was causing it, which team was responsible, or what next month would look like. The FinOps consultants wanted $50K to tell them what a model could compute in seconds. The tools on the market were either too shallow or locked behind contracts sized for companies ten times larger.
So she built the tool herself. Not a spreadsheet, not a monthly report that arrived after the bill closed. A continuous signal: a model that ingested billing data daily, separated deployment-driven spikes from underlying trend, and told the team exactly which resources to address before the charge appeared.
Finopscraft is that tool, built out of the same Dallas office in 2025 and available today for every engineering team that has outgrown AWS Cost Explorer but can't justify a six-figure FinOps retainer.
Our mission
Large enterprises pay hundreds of thousands of dollars for FinOps visibility. They have dedicated cloud finance teams and custom tooling built on top of enterprise agreements. The engineering team at a $10M revenue startup has none of that, but their AWS bill still arrives at the end of the month with no warning and no explanation.
Finopscraft's mission is to give every engineering team the same forecasting capability: per-service, per-region, per-team spend projected 30 to 180 days forward, with specific waste flags and the resource IDs to act on. Starting at $49 per month. Not $50K for a consultant.
We are not building a platform for finance. We are building a signal for engineers. The people who write the code that generates the cloud bill should also be able to see where it is going before it arrives.
The team
Led engineering at two Dallas-area SaaS companies before founding Finopscraft in 2025. Watched cloud costs spiral at both and wrote the first version of the forecasting model in 60 days because no off-the-shelf tool gave her what she needed: a per-service forecast with specific waste flags, not a dashboard full of last month's actuals.
Infrastructure engineer with a background in distributed systems and observability tooling. Designed Finopscraft's billing ingestion pipeline and multi-provider normalisation layer. Believes the best cost tool is one that engineers actually open.
Product background spanning developer tools and API platforms. Focused on making FinOps actionable at the engineering-team level rather than the finance-team level. Runs user interviews, owns the roadmap, and writes the weekly forecast digests before engineers see them.
How we work
The only useful cost signal is one you get before the charge appears. Everything in Finopscraft is designed to surface information early enough to act on it, not to explain what already happened.
We build alerts that land in Jira, not spreadsheets. We label spend by service and tag, not by cost-centre code. Engineers should be able to read a Finopscraft report without a FinOps certification.
Every forecast includes a confidence interval. Every waste flag includes the resource ID, the daily cost, and the specific action to take. We do not recommend "consider right-sizing." We say "this db.r6g.xlarge is running at 4% CPU for 14 days. Downsize or terminate."