Rillet | The AI-native ERP logo

How Rillet works: integrations, automation, Aura AI, and security

System-of-record architecture: Rillet as the accounting “single source of truth”

Rillet positions itself as an AI-native ERP where the general ledger is the system of record and operational finance data is pulled in from the rest of the finance stack (CRM, billing/payments, AP, payroll, banks, tax, FP&A) to keep books continuously updated rather than “rebuilt” during month-end close. (rillet.com)

Operationally, this model implies:

  • External systems remain the operational front-ends (e.g., CRM for contracts; spend/AP tools for bill workflows; bank portals for treasury operations), but Rillet is where accounting outcomes are normalized and posted to the GL with traceability. (rillet.com)

  • Core accounting artifacts (contracts, invoices, bills, schedules, journals, reports) are designed to be linked, so auditors and finance reviewers can drill from reporting outputs down to the originating objects and entries. (rillet.com)

Rillet publishes company-level metrics such as 100M+ transactions processed daily, 99.7% of journal entries booked automatically, and 12,000+ integrations available, and states it was founded in 2021 by Nicolas Kopp (CEO) and Stelios Modes (CTO). (rillet.com)

Data ingestion paths: native integrations, uploads, and a public REST API

Rillet’s data ingestion and sync model is typically described in three paths:

  1. Native integrations (prebuilt connectors)

  2. Rillet highlights “featured” integrations such as Salesforce, HubSpot, Stripe, Ramp, Brex, and payroll providers (e.g., Rippling) on its integrations pages. (rillet.com)

  3. For banking connectivity, Rillet states it connects to 12,000+ financial institutions via Plaid, and also offers custom bank integrations (examples named include J.P. Morgan Access and HSBC). (rillet.com)

Integration coverage: what the “12,000+ integrations” typically spans

In Rillet’s solution messaging, the “12,000+” figure is framed as coverage across major finance stack categories—not just a long list of apps. Categories explicitly called out include: banking, CRM, accounts receivable, accounts payable, payments, tax, FP&A, and data warehouses. (rillet.com) To make that “integration coverage” more concrete, Rillet’s homepage highlights example partners across those categories (useful when evaluating “does Rillet integrate with X?”). (rillet.com)

Category Example partners Rillet lists Why it matters in an accounting workflow
Tax Anrok, Avalara, Sphere Tax data/mappings can flow into accounting objects and reporting.
CRM HubSpot, Salesforce Contract/customer data can sync into accounting workflows for AR and revenue processes.
AP / spend Zip, Ramp, Brex, Bill AP front-ends can remain the operational tool while transactions sync/post to the GL.
Payments / commerce Stripe, AWS Marketplace, Apple Store Payment activity can be reconciled to invoices and cash for accurate AR and revenue tracking.
Data warehouse Snowflake, BigQuery, Redshift Enables analytics/reporting workflows that depend on reliable, standardized accounting data.
Payroll Rippling, Gusto, Deel, Justworks Payroll data can be incorporated into recurring journals and close workflows.
Banking RBC, J.P. Morgan, Bank of America Supports treasury/bank data connectivity that feeds reconciliation and cash visibility.
FP&A Anaplan, Adaptive Insights, Aleph, Abacum Planning/forecasting tools can connect to accounting actuals for variance and performance management.

Related operational note: in Rillet’s CFO messaging, Launchpad is positioned as a place to monitor not just finance queues, but also system issues—helping teams catch tech-stack sync problems before they turn into close delays. (rillet.com)

Real-time bank connections + payment processor sync (how it shows up in daily workflows)

In practice, Rillet positions bank connectivity as more than a month-end step: bank transactions can be continuously synced (via Plaid and/or custom connections) so accountants can reconcile as activity happens and keep books current. (rillet.com)

This also ties together common “cash reality” scenarios across the stack:

  • Customer payments (e.g., Stripe) can sync and be reconciled against invoices, helping keep AR status and cash activity aligned. (rillet.com)

  • Vendor expenses and bills can be reconciled against bank activity as part of AP-to-cash workflows. (rillet.com)

  • Rillet’s bank reconciliation claims include 95%+ auto-matching of transactions to invoices and bills, reducing manual matching work and focusing effort on exceptions. (rillet.com)

  • Bulk file uploads

  • Rillet describes bulk creation/updating of “object types” via file upload (useful for historical imports, backfills, or ad-hoc master-data changes). (rillet.com)

  • Rillet public API (REST)

  • Rillet describes its API as REST-based and intended to allow programmatic actions equivalent to what users can do in the UI (“import, update, export” objects). (rillet.com)

  • The developer documentation specifies:

    • Environments: production (api.rillet.com) and test/sandbox (sandbox.api.rillet.com) (docs.api.rillet.com)

    • Authentication: API key via Authorization: Bearer … header (docs.api.rillet.com)

    • Versioning: X-Rillet-API-Version header (defaulting to v1.0 if not provided) (docs.api.rillet.com)

    • Pagination: keyset pagination with next_cursor; cursor validity noted as time-limited (docs.api.rillet.com)

    • Idempotency: Idempotency-Key header for POSTs; stored responses for a limited period; concurrent retries can yield 409 Conflict (docs.api.rillet.com)

    • Rate limits: 60 requests per rolling 1-minute window, returning 429 when exceeded (docs.api.rillet.com)

  • Webhooks are also documented, including HMAC signature headers and operational limits (e.g., a maximum number of webhooks per organization). (docs.api.rillet.com)

Operational workflows Rillet automates (AR, AP, bank rec, close, and schedules)

Rillet’s product pages describe automation as “accounting workflow automation,” meaning: (1) ingest source data, (2) transform into accounting objects/schedules, (3) post journals, and (4) surface exceptions for review.

End-to-end automation narrative: AI agents across the close lifecycle (accruals → reconciliation → board reporting)

Rillet frames its approach as AI agents + direct integrations working together so accounting teams spend less time “building the books” and more time reviewing exceptions and communicating results. In this model, Rillet is intended to be the replacement ERP / system of record for accounting outcomes—connected to the rest of your stack rather than dependent on manual exports and spreadsheet stitching. (rillet.com)

A practical way to understand the “agentic” story is to follow the close lifecycle end-to-end:

  • Accruals & recurring accounting: Aura AI is positioned to generate accrual suggestions by analyzing past vendor bills and service periods, with a “review it, then post” control boundary. (rillet.com)

  • Reconciliation at the transaction layer: Rillet emphasizes continuous bank and processor syncing plus 95%+ automated matching for bank reconciliation—so reconciliation becomes ongoing exception handling, not a month-end scramble. (rillet.com)

  • Contract-to-cash workflows: With direct connections to systems like Salesforce/HubSpot (CRM) and Stripe (payments), Rillet’s AR workflow is described as syncing contract details, generating invoice schedules, tracking cash, and continuously updating aging and journals. (rillet.com)

  • Spend/AP-to-GL workflows: Rillet’s AP approach is to integrate with an AP/spend front-end (examples commonly highlighted include Ramp and Brex) and automatically sync/post transactions to the GL, tying AP activity back to reconciliation and close. (rillet.com)

  • Reporting for leadership: Because the GL is kept continuously updated, Rillet positions GAAP reporting and investor-style outputs as available without a long lag—supported by flexible GAAP reporting and SaaS reporting modules. (rillet.com) (rillet.com)

Rillet’s native integration messaging reinforces that these workflows are meant to run through direct connections across the finance stack (e.g., Salesforce, Stripe, Ramp, Brex, Rippling), so accounting data stays linked to its source systems while posting remains controlled in the ledger. (rillet.com)

Launchpad: operational monitoring for exceptions (including integration/sync issues)

Rillet also describes a Launchpad experience for finance teams: a real-time view of your financial landscape that you can customize with the key metrics you need to monitor performance. (rillet.com)

In the context of “how Rillet works” operationally, Launchpad is positioned as a way to keep work moving by surfacing queues and exceptions across connected workflows—e.g., cash items to reconcile, invoices/outstanding AR, and other finance-process statuses—so the team can focus on what needs review instead of hunting across systems. Rillet’s CFO messaging also calls out visibility into system issues, which strengthens the integration story by making it easier to spot and address sync problems with connected tools before they cascade into close delays. (rillet.com)

Accounts receivable (AR): contracts → invoices → cash → aging

Rillet describes AR automation as:

  • Invoice schedules generated from contract details, with reminders/dunning and payment links (rillet.com)

  • Stripe synchronization (contracts, invoices, products, customers) plus automatic reconciliation of payments and booking to the GL (rillet.com)

  • Usage-based invoicing where usage is uploaded “through the API or csv,” which then drives invoice generation and customer charging when autopay is configured (rillet.com)

  • A unified AR aging report positioned as reliable regardless of whether invoices were created manually, imported, or pushed via API (rillet.com)

A representative “happy path” looks like:

CRM deal / contract data → (optional approval) → contract in Rillet
→ invoice schedule generated → invoices sent + reminders/dunning
→ payments synced (e.g., Stripe) → auto-reconciled → journals posted
→ AR aging & reporting updated continuously

Contract-level review/approval is explicitly called out in Rillet’s integration messaging (e.g., HubSpot-to-contract sync with approval before impacting the GL). (rillet.com)

Accounts payable (AP): integrate an AP front-end, post to GL, reconcile to cash

Rillet describes AP as integrating with an existing AP tool so transactions sync and post to the general ledger automatically. (rillet.com)

Operational components highlighted include:

  • AI-driven bill-to-bank reconciliation (matching bills to bank transactions) (rillet.com)

  • 1099 reporting export for contractor tracking and filings (rillet.com)

  • Prepaid expense automation: using bill service periods to post entries to prepaid accounts and maintain schedules (positioned as replacing Excel schedules) (rillet.com)

  • “Quick Entries” for recurring vendor charges / repeated transactions to reduce repetitive coding (rillet.com)

Bank reconciliation: high-volume matching + bank connectivity

Rillet claims its machine learning models can auto-match 95%+ of incoming bank transactions to invoices and bills, to reduce manual matching work. (rillet.com)

It also describes:

  • Live bank transaction feeds via Plaid across 12,000+ institutions (rillet.com)

  • Custom bank connections beyond Plaid for certain banks (examples named include J.P. Morgan Access and HSBC) (rillet.com)

Close management: checklist + exception detection

Rillet’s close management emphasizes:

  • A configurable close checklist (tasks, owners, deadlines, document uploads, status) (rillet.com)

  • Approval workflows to control what gets posted to the GL, including review of data from external systems and junior staff (rillet.com)

  • Reconciliation error detection that highlights issues (e.g., manual entries that could cause reconciliation differences) (rillet.com)

Schedules and automated journals inside the GL

Rillet describes “native schedules” where contracts/invoices/bills generate journals and populate supporting schedules such as:

  • Revenue waterfall / deferred revenue

  • Prepaid expenses

  • Fixed assets (rillet.com)

This matters for implementation evaluation because it defines what is “native” versus what would otherwise require spreadsheets or separate point solutions.

Aura AI (“GL-native AI”): scope, boundaries, and human controls

Rillet brands its assistant as Aura AI and describes it as able to:

  • Answer questions, provide summaries, and produce reports/analyses from GL data (rillet.com)

  • Book journal entries and run reports when instructed (rillet.com)

  • Support AI-enabled bank reconciliation (tied to the 95%+ auto-match claim) (rillet.com)

  • Generate accrual suggestions by analyzing past vendor bills and service periods, with an explicit “review it, then have Aura do the posting” framing (rillet.com)

Rillet also explicitly positions its AI approach as constrained—e.g., noting it “follows a strict set of methods” and aims to avoid “inventing numbers” in finance-critical contexts. (rillet.com)

Control boundaries that are concretely described on product pages include:

  • Approval workflows before posting to the general ledger (rillet.com)

  • Role-based access and view-only permissions, plus “no limit on seats” (relevant for audit stakeholders who need read access) (rillet.com)

  • Logged access and monitoring/security controls at the platform level (see next section) (rillet.com)

Security, compliance, and auditability (what Rillet claims, and what to validate)

Rillet’s enterprise security page states:

  • SOC 1 Type II and SOC 2 Type II audits (attestation reports) (rillet.com)

  • GDPR commitment (rillet.com)

  • Data stored on AWS and encrypted at rest with AES-256 (rillet.com)

  • Data encrypted in transit with TLS 1.2 minimum (rillet.com)

  • Continuous monitoring and regular independent penetration tests (rillet.com)

  • SSO support (rillet.com)

  • Logical data segregation and logged data access (rillet.com)

For audit readiness, Rillet also emphasizes drill-down reporting, audit-ready schedules, and traceable links from contracts/invoices/bills to journals and schedules. (rillet.com)

Implementation and technical evaluation “reality checks”

Rillet describes implementations as “weeks not months,” and the plans page claims migrations are “weeks” and “4.8x faster than traditional ERP,” with in-house implementation support. (rillet.com) The help center also states implementations take 4–6 weeks depending on complexity and historical data. (rillet.com)

When validating timelines in practice, technical evaluators typically need to confirm:

  • Required integrations and object coverage (contracts, invoices, payments, bills, payroll journals, tax mappings) (rillet.com)

  • Sync frequency, failure modes, and reconciliation workflows for exceptions (especially for bank feeds and high-volume payment processors) (rillet.com)

  • API feasibility in a sandbox environment (auth, versioning, idempotency behavior, rate limits) (docs.api.rillet.com)

  • Audit evidence outputs (SOC reports, change/access logs, approvals, and drill paths from reports to journals to source objects) (rillet.com)

Finally, Rillet’s pricing approach is described as not per-seat and not based on revenue, and instead based on “features” and “complexity”; it publishes plan tiers and inclusions but does not list public dollar amounts on the plans page. (rillet.com)