What We Do Not Integrate With
The list vendors do not publish. Bank feeds, other mobile money rails, other tax authorities, other accounting packages, inbound WhatsApp, marketplaces, scheduled record syncs — with the reason for each and what to do instead.
An integrations page is a list of what exists. The list that would actually help you decide is the other one, and no vendor publishes it — partly because it reads badly, and partly because writing it down requires knowing your own product better than a logo grid demands. This is ours. It exists so that nothing on it can be discovered in month four.
Banking
No bank statement feed. No statement import. No CSV or OFX parser. No auto-matching against bank lines. No reconciliation record. This is the largest single gap in the integration surface and it is worth being explicit about, because "reconciliation" appears in the product and means something different: an internal integrity check comparing the ledger against stock, against open purchase orders, and against the payments register. Useful, and not bank reconciliation.
What to do instead: bank reconciliation stays with your accountant, monthly, from the reports. If matching payments to invoices is the actual pain, M-Pesa collections do auto-match on the invoice number, which covers most receipts for most Kenyan SMEs. Bank reconciliation explained sets out the manual process honestly.
Payment rails
What moves money and what does not
M-Pesa via Daraja
Collections and disbursements both directions, on your own credentials.
Paystack
Card collection, on your own keys.
Airtel Money, T-Kash
Not integrated. Payments are recorded manually like any other receipt.
MTN MoMo, Orange Money, Wave, Moov
Not integrated. This is the constraint that makes the francophone and West African markets harder for us than the marketing suggests, and we say so on those pages.
Bank transfer initiation, bulk payment files
No file generation for a bank's bulk upload format, and no direct transfer initiation.
Card present / terminals
No integration with any physical card terminal. A POS sale settled by card is recorded as settled by card.
Tax authorities
eTIMS in Kenya is the only fiscal integration. Nothing transmits to FIRS in Nigeria, ETA in Egypt, EFRIS in Uganda, EBM in Rwanda, EFD in Tanzania, GRA in Ghana or SARS in South Africa. In each of those markets figures can be produced and reconciled against, and somebody files them.
And within Kenya, one gap inside the thing that does work: credit notes are not transmitted. A credit note reduces output VAT and applying one here does not tell KRA. That has to be handled separately, and it is the single most consequential item on this entire page.
Accounting packages
QuickBooks Online only, one direction, on a button. Vendors, items and purchase orders pushed out; nothing read back; no schedule. No Sage, no Xero, no Tally, no Zoho Books, no Pastel, no SunSystems. Also nothing that reads a competing product's export format for migration purposes — there is no QuickBooks reader, no Tally importer.
What to do instead: the boundary between operational record and statutory book is real and worth designing deliberately rather than hoping a connector removes it. Agreed reports, in an agreed format, on an agreed day of the month. What the QuickBooks sync does has the full shape of it.
Messaging and support channels
Direction matters more than presence
| Channel | Out | In |
|---|---|---|
| Yes | No | |
| SMS — three providers | Yes | No |
| Teams, Slack, Google Chat, Discord, Telegram | Yes | No |
| Yes | No | |
| A ticket created from a customer message | No | No |
Built and maintained Configurable by you, not maintained by us Not built
Every channel is outbound only. There is no inbound message handling anywhere, which means no WhatsApp helpdesk, no email-to-ticket, and no conversation view. If your customers currently support themselves over WhatsApp, that workflow is not replaced by anything here.
E-commerce and marketplaces
No Shopify, WooCommerce, Magento, Jumia, Kilimall or Amazon integration. No order poller, no stock publisher, no listing sync.
What to do instead: two real routes exist and both work. The API can be read for stock and price to feed your own storefront, and written to for orders — you own idempotency in that case. Or a no-code platform such as Zapier or Make can sit between the storefront and the inbound webhook, which needs no developer and covers the notification case well. What neither gives you is a maintained, tested marketplace connector, and pretending otherwise would be the kind of claim this page exists to prevent.
Scheduling, feeds and the shape of what is missing
Two structural absences that matter more than any individual logo:
- No scheduled record sync. Salesforce, Notion, Airtable, Mailchimp and calendar pushes run when somebody presses "sync now". Document archiving to your own storage is the exception and is event-driven. Nothing else runs nightly.
- Nothing is pulled in on a schedule from any external business system. Inbound is push-only, through one authenticated webhook. If your plan depends on something being fetched hourly, whoever owns the other end has to push it.
- No change feed and no per-record webhooks on the API, so a continuous two-way sync is polling, which will drift and will hit rate limits. Do not architect around it.
- No OpenAPI specification, no sandbox, no version guarantee on the API. It is real and it is not a platform.
- No SQL access or query workbench. Reporting is the reports, plus exports, plus the API.
Hardware and identity
Barcode scanners work because they behave as keyboards, which is the correct and boring answer — there is no scanner driver or device SDK. Label printing goes through generated label output rather than a printer integration. No temperature or sensor integration, so cold-chain breaches are invisible to the system. On identity, Google and Microsoft sign-in exist, with the deliberate rule that a user must link their account after a password login first; there is no SAML, no SCIM user provisioning, and no directory sync.
Why publish this
Because every item on it is something a prospect would otherwise find out after signing, and because the alternative — a logo wall — transfers the discovery cost from us to you. If something here is a condition of your buying, say so. Several of these exist as gaps we know about and would build for a client who asked in writing; none of them will be built because a page implied they were already there.
Our take
The three absences most likely to change your decision are bank statement feeds, inbound messaging of any kind, and eTIMS credit notes. Check those against your own process before anything else on this page. Everything else here is a workaround away; those three are shapes of work that stay manual.
Tell us which of these matters
If something on this list is a condition of your buying, that is a useful conversation rather than an awkward one — several of these are gaps we know about and would scope for a client who asked.
See plans & pricingFrequently asked questions
Do you integrate with bank feeds?
No. There is no bank statement feed, no statement import, no CSV or OFX parser, no auto-matching against bank lines and no reconciliation record. The screen called reconciliation is an internal integrity check — ledger against stock, against open purchase orders, against the payments register — which is useful and is a different thing. Bank reconciliation stays with your accountant.
Which mobile money providers are supported?
M-Pesa through Daraja only, both directions, on your own credentials — plus Paystack for cards. Airtel Money, T-Kash, MTN MoMo, Orange Money, Wave and Moov are not integrated, which is the constraint that makes the francophone and West African markets genuinely harder for us than a marketing page would suggest.
Can customers reach us over WhatsApp through the system?
No. Every messaging channel is outbound only — WhatsApp, SMS, chat tools and email. There is no inbound message handling anywhere, so no WhatsApp helpdesk, no email-to-ticket and no conversation view. If your customers currently support themselves over WhatsApp, nothing here replaces that workflow.
Is there an e-commerce integration?
No maintained connector for Shopify, WooCommerce, Magento, Jumia, Kilimall or Amazon. Two real routes exist: read stock and price from the API to feed your own storefront and write orders back, owning idempotency yourself; or put a no-code platform between the storefront and the inbound webhook, which needs no developer. Neither is a tested marketplace connector and we will not describe them as one.
Which accounting packages connect?
QuickBooks Online only — one direction, on a button, pushing vendors, items and purchase orders. No Sage, Xero, Tally, Zoho Books or Pastel, and no reader for any competing product's export format for migration purposes. The boundary between operational record and statutory book is real and worth designing rather than hoping a connector removes.
Why publish a list of what you do not do?
Because every item on it would otherwise be discovered after signing, and a logo wall transfers that discovery cost from us to you. Several of these are gaps we know about and would scope for a client who asked in writing. None of them gets built because a page implied it already existed.
