Multilingual AI client intake

Your client speaks Arabic.
Your caseworker reads English.
No interpreter needed at intake.

Legal clients who don’t speak English well face two bad options: struggle through an English intake and lose the details that decide a case, or wait days for an interpreter while the matter stalls. Neither serves the client or the firm.

See it work See the languages
How it works

The client speaks their language. Your team reads English.

One intake, two languages — and the privacy step sits in the middle, before any AI is involved.

1

They tell their story

The client opens the intake on your website and answers in their own language — plain questions, one at a time, no forms.

2

The AI listens in that language

The assistant runs the whole intake conversation in the client’s language, checking the answers make sense and filling the gaps.

3

Identifiers are stripped first

Before any AI processing, names, addresses and identifiers are removed server-side — the model never sees who the client is.

4

An English case file arrives

Your caseworker or lawyer opens a structured English case file — facts, timeline, parties, what’s missing — ready for the first appointment.

60+ languages

Built for the languages your clients actually speak.

The client picks their language at the start. Whatever they choose, your team receives the file in English.

العربية
Arabic
中文
Mandarin Chinese
Español
Spanish
Français
French
Kreyòl
Haitian Creole
Tiếng Việt
Vietnamese
Soomaali
Somali
دری
Dari
پښتو
Pashto
Українська
Ukrainian
Русский
Russian
हिन्दी
Hindi
বাংলা
Bengali
Ελληνικά
Greek
Türkçe
Turkish
한국어
Korean
日本語
Japanese
Português
Portuguese
Italiano
Italian
Bahasa
Indonesian
+ 40 more
and counting
Who it’s for

Two kinds of organisation, one multilingual problem.

Community legal centres & legal aid

You serve culturally and linguistically diverse (CALD) communities, but interpreter availability throttles your intake and details get lost at first contact. Clients intake in their own language; your caseworker opens a structured English case file — no interpreter needed just to find out what someone needs.

For community legal centres

Immigration & personal injury firms

Your highest-volume clients often don’t speak English well, and an interpreter at the first call is slow and costly — so good matters slip away. The client tells their story in their language; a qualified English matter file is in your inbox before the first call.

For law firms Personal injury intake Immigration intake

Intake in any language is exactly as private as intake in English. The client’s personal details are stripped on our server before any AI sees them, the data is processed transiently and permanently deleted on delivery, and it runs on your own AI key. A client intaking in Somali is as protected as one intaking in English. See the full architecture →

See it work

One picker — then everything is in their language.

Community Legal CentreAI legal intake Secure

Choose your language

Search languages…
English
中文Chinese (Mandarin)
粵語Chinese (Cantonese)
EspañolSpanish
العربيةArabic
हिन्दीHindi
PortuguêsPortuguese
كيف يمكننا مساعدتك اليوم؟
فُصلت من العمل دون إشعار.
الوقائع المُسجَّلة
نوع المسألةفصل تعسفي
صاحب العملمُسجَّل
الإطار الزمنيمكتمل

The client selects their language at the start of intake. Everything from that point — questions, answers, document prompts — runs in that language.

Set it up for your organisation
Questions

Multilingual intake, answered.

Which languages does PROVEAiBLE support?
60+ languages, including Arabic, Mandarin Chinese, Spanish, French, Haitian Creole, Vietnamese, Somali, Dari, Pashto, Ukrainian, Russian, Hindi, Bengali, Greek, Turkish, Korean, Japanese, Portuguese, Italian and Indonesian. The client chooses their language at the start of the intake. Whatever language they use, the case file your team receives is in English. The models behind it are Claude and ChatGPT.
Does the client need to type, or can they speak?
It’s a written conversation — the client types their answers in their own language, one plain question at a time, with no forms and no jargon. On a phone they can use their device’s built-in voice-to-text if they prefer, but nothing special is required to set up.
What language does the case file arrive in?
Always English. The client tells their story in their own language; your caseworker or lawyer receives a structured English case file — facts, timeline, parties and what’s missing — ready for the first appointment. The client also gets a plain-language summary of what they shared, in their own language.
Is intake in other languages as private as English intake?
Yes — the privacy architecture is identical regardless of language. Personal identifiers are stripped on our server before any AI model is called, the data is processed transiently and permanently deleted on delivery, and it runs on your own AI key (BYOK). A client intaking in Somali is exactly as protected as one intaking in English. See the architecture →
Can we configure which languages appear for our clients?
Yes. We set the intake up around your client base during setup — you can prioritise the languages your community actually speaks, or keep the full set available. Every language uses the same redaction and English case-file pipeline; configuring the options doesn’t change the privacy or the output.
Pilot it at your organisation