SDK TypeScript · apps & products

Build the tool.
ASTRIA does the math.

Internal apps, client portals, SaaS products: embed projections and Monte Carlo in your UIs with a typed client — without coding an engine or managing infrastructure.

$npm install @subspacecomputing/sdk

Node.js 18+ · npm · v0.1.1 · aligned with the REST API · npm package

calculator.tsready● @subspacecomputing/sdk
// API responses
your applicationpreview · ASTRIA

Savings calculator

—projected median capital · 12 years · start $10,000

p10 — · p90 — · 1,000 scenarios

Powered by ASTRIArun — · — ms

demo — pick a profile: the code updates and the run replays. Same profile, same fingerprint, every time.

From internal widget to billed product.

The same TypeScript client, from Notion-replacement prototype to production SaaS.

Internal tools

Scenario simulators for leadership, price calculators for sales, dashboards that recalculate when an assumption changes.

Client portals

Different projections per client, computed on demand under your brand — with no engine living in your code.

SaaS products

Ship computation as a billed feature of your product. Batch by entity with projectBatch() on Business plans and above.

Two more products, the same client.

A projection table for your analysts, a mortgage renewal for your advisors — each is just a project() call behind your interface.

Savings projectionready

—projected balance at 8 years

contributions —
interest —

$6,000/yr · return — · annual compounding

YearContributionsInterestBalance

contributionsinterest — of balance

✓ deterministic · project() · outside quotarun —

Mortgage renewalready
balance —remaining —maturing rate —

——

Offered termRatePaymentBalance at term end
Stress test · qualifying rate✓ qualifies
qualifying rate—qualifying payment—GDS—limit39.0%

qualifying: the greater of 5.25% or rate + 2% · project()

✓ 3 rate scenarios · a single call · —run —

Three calls, and you're in production.

Only simulate() consumes your quota — prototype your UI with project(), then move to Monte Carlo when you're ready.

01

client.project(spec)

Deterministic 1-scenario projection to wire your UI at no cost.

outside quota

02

client.simulate(spec)

Multi-scenario Monte Carlo: full statistics and run_id.

quota · full stats

03

client.getRun(id)

Status, summary, and artifacts for every run — audit built into your app.

auditable · traceable

The details that matter at 2 a.m.

Typed end to end

Typed responses and errors — QuotaExceededError, RateLimitError, ValidationError with stable errorReason . Your IDE knows the shape of every response.

The key stays on the backend

Built for Node or a controlled proxy: your be_live_ key never reaches the browser. Multi-team context via setTeamId().

Visible quotas and limits

getUsage(), getQuotaInfo(), getRateLimitInfo(): your app knows where it stands before your customers find out.

Your first widget calculates this afternoon.

from $49/mo · 25,000 scenarios included · project() outside quota for prototyping