AI Assistants CRM Email + Projects Live Chat Ticketing Flows Enterprise Security HR-lite Control Center Solutions Pricing Start free trial
For accountancy practices

Run the practice.
Protect the numbers.

From onboarding and AML checks to engagement letters, billing and the year-end scramble — one governed workspace where the AI helps with everything but never sees a client’s raw financials in the clear.

Client onboarding
AML check · passed
Engagement letter · sent for e-sign
client_4821 🔒

Built for how accountants work.

1

Onboard & AML

Capture clients via Flows, run checks, store evidence governed.

2

Engage & e-sign

Generate the engagement letter, send for signature, file it.

3

Bill & collect

Quote → invoice → pay in the CRM, chased automatically.

4

Advise with AI

Draft, summarise and analyse in Workbook — on tokenised data.

Confidentiality by construction

Client financials the AI
never reads in the clear.

Names, turnover and account numbers are tokenised before any model. You keep the AI productivity; your clients keep their confidentiality; your file keeps the audit trail — helpful when AML or PII queries land.

Governance & DSAR →
🔒

Tokenised financials

Turnover, UTR and account numbers masked before AI.

📋

Evidence on file

AML & engagement records captured and audited.

Client isolation

Matters keep one client’s data away from another’s.

The engagement lifecycle

One governed thread from prospect to year-end.

A client doesn’t arrive as a tidy folder. Other Me carries them through the whole engagement — capture, checks, letter, work, billing, advice — without your AI ever touching raw financials in the clear.

1

Enquiry & capture

A prospect fills in a Flows form or books a discovery call through a scheduling page. The enquiry lands as a CRM lead — no rekeying, nothing lost in an inbox.

2

AML & onboarding checks

Convert the lead to a client record, run your AML/KYC checks, and store the evidence against the file. The audit trail records who did what and when.

3

Engagement letter & e-sign

Draft the engagement letter in a Document, send it for built-in e-signature, and file the executed copy back against the client — all inside one workspace.

4

Records, year-end & advice

Pull schedules together in Workbook, draft the covering letter, and ask the AI to summarise or explain — on tokenised figures, never the client’s real numbers.

5

Bill, collect & retain

Quote → invoice → pay via a public Revolut pay page, with recurring retainers for monthly clients and automatic chasing on what’s overdue.

Henderson Ltd — FY year-end on track
AML CHECK
Passed
ENGAGEMENT
Signed
RECORDS
In review
FEE NOTE
Drafted
What the AI sees on this file:
Henderson Ltdclient_4821🔒
Tokenised before AI
Onboarding & AML

Take on clients without losing the paper trail.

UK practices carry anti-money-laundering and customer due-diligence obligations under the supervision of bodies like the ICAEW — which means every new client needs identity checks, risk assessment, and evidence you can produce later. Other Me keeps that evidence where the engagement lives.

  • Capture enquiries through Flows forms and booking pages — straight into a CRM lead.
  • Convert to a client record and attach your AML/KYC evidence against the file.
  • Every action is recorded in the audit trail — who, what, when, on which client.
  • Offboarding triggers the SCRS leaver kill-switch, logically revoking data access.

Context, not a claim: ICAEW/AML duties are your obligations. Other Me is a governed workspace that helps you meet them — it is not a certification or a supervisory body.

New client · due diligence
Identity verified · passed
Risk rating · standard
Evidence filed · 2 documents
audit trail 🔒
Documents & year-end

Schedules, letters and analysis — drafted, not leaked.

Year-end is where confidential figures and AI usually collide. With Other Me you draft covering letters in a Document, build and reconcile schedules in Workbook, and ask the AI to summarise or explain — while real names, turnover and account numbers are tokenised before any model sees them.

  • Free Document, Workbook and Deck editors with every AI plan — no add-on.
  • Best-model-per-job routing across GPT, Claude, Gemini and Grok — no per-tier gating.
  • The model sees client_4821; you see Henderson Ltd — reversible tokenisation.
  • Nothing readable sits in the search index; content stays in a separate encrypted store.
Documents & Workbook →
AI prompt on a live file

“Summarise the movement in trade debtors and draft a note for the file.”

Henderson Ltdclient_4821🔒
£482,000amount_1🔒

Real figures are masked on the way out and restored in your view on the way back.

How the firewall works

Why your client’s numbers never reach the model.

SCRS is a firewall, not a filter — it’s patent-pending (UK application 2602911.6). Out-of-scope data is removed before a result exists, and anything sensitive is tokenised before a third-party model is ever called.

Block before search

Permissions are enforced during retrieval, so out-of-scope client data is removed before any result is formed — pre-retrieval security, not a post-hoc scrub.

🔒

No plaintext in the index

The search index holds no readable text. Content sits in a separate encrypted store and is released only after cryptographic verification — fail-closed, so any failure returns nothing.

Tokenised before the LLM

Eleven built-in regex types (National Insurance, UK postcodes, payment cards, dates of birth, passport numbers and more) plus name/organisation detection mask sensitive data before any model call.

UK/EU data residency No client data used to train models Patent-pending firewall See the security model →

Workflows a practice runs every week.

Self-assessment season

Track every return on a shared Kanban board, set reminders before deadlines, and chase missing records through Email and Flows — without dropping a client.

Monthly bookkeeping clients

Recurring retainers invoice themselves; the revenue dashboard shows what’s billed and outstanding so cashflow never depends on someone remembering.

Year-end accounts

Build schedules in Workbook, draft the covering letter in a Document, and have the AI summarise movements — all on tokenised figures.

New client onboarding

Enquiry form → lead → AML checks → engagement letter with e-sign → first invoice, with the evidence filed and audited end to end.

Client queries & support

Run a shared inbox with SLAs and a knowledge base in Ticketing, or a governed Live Chat agent on your site that turns enquiries into leads.

Advisory & planning

Generate scenario decks and explainer notes for clients, drafted with AI and grounded in your own governed data — never trained on it.

“The pitch most accountants fear is ‘put your clients’ figures into an AI.’ Other Me flips it: the model only ever sees tokens, the real numbers stay in an encrypted store, and there’s an audit trail when a client asks who touched their data.”

Illustrative scenario — not a named client.

Honest about where we are

What’s here now — and what’s coming.

Available today
  • UK GDPR compliant; UK Age Appropriate Design Code compliant.
  • SCRS firewall, tokenisation and audit trail across the workspace.
  • CRM, Email + Projects, Documents, Flows, Ticketing, Live Chat, HR-lite.
  • Built-in e-sign and Revolut pay pages for fee notes and retainers.
On the roadmap
  • ISO 42001 — on track (AI management system).
  • SOC 2 Type II and ISO 27001 — on the roadmap (Q4 2026 / 2027).
  • Enterprise — SSO, BYOK, data-residency controls and matter walls, coming soon.
  • HIPAA available by BAA/contract where it applies.

Nothing here is certified yet — we say “on track” / “on roadmap” and mean it. Talk to us about Enterprise early access.

The stack for a practice

AI base (with free editors) + CRM for clients & billing + Email + Projects. Flows & HR-lite come free with CRM.

AI · £24CRM · £12Email + Projects · £8Flows free
Start free →
Governed AI for accountants

Give the practice an AI it can
actually put on client files.

Onboard, engage, bill and advise in one workspace — with the numbers tokenised before any model and an audit trail when anyone asks.

Already with us? Sign in.