Commentary at midnight
The numbers are done. Now someone has to explain every variance in plain English before the deadline.
For finance and fintech teams
Close faster, explain everything
Variance commentary, invoice and statement extraction, reconciliation notes and board pack drafts. Every call is logged with tokens and price, and you decide exactly what data is sent.
Faster close
commentary drafted from the numbers
Every call logged
tokens, price and model on record
Your data, your call
only what you send is processed
The problem
What is slowing you down today
Documents that will not parse
Invoices, statements and remittances arrive in a hundred layouts, and rules-based extraction breaks on every new one.
Tools nobody can audit
When AI is used in a finance process, you need to show what was asked, which model answered and what it cost.
Capabilities
Everything you need, nothing you don't
Variance commentary drafts
Send actuals, budget and context, get clear explanations for the analyst to check and sign off.
Document field extraction
Ask for vendor, dates, amounts and line items as JSON from the text of an invoice or statement.
Full request history
Model, tokens, latency, price and a short preview recorded for every call.
Hashed keys and rate limits
Keys are shown once and stored as a hash. Per-account rate limits keep automations in check.
Reasoning where it counts
Route analysis to a reasoning-capable model and simple formatting to a lighter one.
Idempotent batch runs
Send an Idempotency-Key so a retried job is not metered twice.
Use cases
What teams run on it
01
Month-end variance notes
First-draft commentary on every material line for review.
02
AP document intake
Structured fields from invoice text, ready for your matching workflow.
03
Policy and procedure Q&A
Answers to expense and approval questions from the policy text you supply.
04
Board pack narrative
Turn the KPI table into a readable summary for directors.
How it works
From sign-up to production
- 1
Prove it on one report
Run last month's variance table through the playground.
- 2
Integrate with controls
Use dedicated keys for each workflow and set rate limits in settings.
- 3
Review your obligations
Confirm what data your policies allow you to send, then scale.
Control
Built so you stay in charge
Keys you control
API keys are shown once at creation and stored only as a hash. Revoke any key instantly from Settings.
Spend you can see
Every request carries its own receipt: tokens in, tokens out, latency and price. Nothing is billed that you cannot trace.
Limits you set
Per-account rate limits and a free monthly allowance keep experiments safe before they become production.
Data you decide on
You choose exactly what you send. Repeat prompts can be served from your cache, and cache hits are never billed.
For developers
One endpoint. Every model.
Send a prompt and a capability, or name a model. One bearer key, one small JSON body, and a full receipt on every response.
curl https://YOUR-DOMAIN/api/ainf/inference \
-H "Authorization: Bearer $AI_INFERENCE_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: order-1042" \
-d '{"prompt":"Summarise this ticket in one line","capability":"chat"}'Questions, answered
Is this suitable for regulated financial data?
You control exactly what each prompt contains. Review your own regulatory and contractual obligations before sending sensitive or personal data. We make no compliance claims on your behalf.
Can we audit what was sent?
Request history keeps the model, tokens, price, latency and a short preview of each call.
Will a retried batch double-bill us?
Not with an Idempotency-Key. Usage for a key is metered once.
How is pricing shown?
Per-token prices are listed for every model before you call, and each response reports its own price.
Make month-end a little less heroic
Start on the free allowance with one report. Keep every call on the record.