Precomputing: Own Your Analytics Without Paying Per-Gigabyte to a Third Party
Bootstrapped founders know that every line item in your SaaS bill cuts into profit. The hosted analytics service you signed up for looked cheap at first, but now you’re paying per gigabyte ingested, per query, per custom metric. You’ve got months of customer data, but you hesitate before running a new analysis because you know it will cost money. Meanwhile, you still don’t have the dashboards you actually need.
The other approach is to build analytics from scratch. You recalculate numbers on every dashboard refresh, which is fine until you get actual traffic. Then every refresh becomes slow, and every slow dashboard is a feature you’re not shipping because you’re spinning waiting for a query.
Precomputing flips the model. You calculate answers once, as the data comes in. You write a short policy that names the answers you want: requests per endpoint, usage by customer, error rates by type. You say how long each level of detail should live. Precomputing compiles that policy to SQLite triggers, so every INSERT keeps those answers fresh, and reading an answer is a single lookup.
For bootstrapped companies, the payoff is immediate: you own your data and you pay nothing for analytics. The data stays on your machine. There’s no per-query fee, no per-gigabyte ingestion cost, no vendor lock-in. When you want to know how many requests hit the API last week, you query your own database and get the answer in milliseconds.
The architecture is simple enough that you can maintain it alone. The policy language is readable: it names answers and says how long they should live. Precomputing compiles to SQLite triggers, which run invisibly as data arrives. Your dashboard queries views, same as any other SQL. When triggers get slow, you can swap in a Go engine that runs the same policy, writes the same file, and your dashboard never knows it happened.
The same policy drives other things bootstrapped companies need. A usage meter for billing where a retried request counts once and a closed month stays locked. You don’t lose money to billing edge cases. A log reducer that keeps every line locally but sends upstream only what you really need, so your logs don’t bloat your upstream service costs. You stay in control.
Precomputing is version 0.1, tested hard on simulated data and not yet run on production traffic. The platform page covers the architecture. The Policy Language walks through a policy line by line. The SQL demo puts three hours of simulated traffic through a compiled policy in SQLite’s WebAssembly build, right in your browser.
For bootstrapped founders, the value is sovereignty and economics. You own your analytics infrastructure. You pay nothing for analytics. Your data stays on your machine. You understand how your metrics are calculated and you can change them whenever you want. When you have a question about your business, you ask your own database, not an API that charges per query.
That’s how bootstrapped companies stay profitable: they keep data where it belongs, on infrastructure they control, and they use tools that let them ask their own questions for free. Precomputing is built for founders who want to own their analytics stack.