TALKLYby Curiosity Technology

Real-time voice translation

Speak in your language. They hear it in theirs.

Two people. Two languages. One conversation.

TALKLY hears the conversation, understands it and carries it to the other person in their own language.

The problem

Most conversations don’t fail. They stall.

A shared language is assumed in more places than it exists.

  1. 01 · Reception

    Directions are given twice, understood once, and the guest still takes the wrong lift.

  2. 02 · Meeting

    The discussion stops while someone looks for the right word in a language that isn’t theirs.

  3. 03 · Clinic

    A patient repeats the same sentence three times, a little louder each time.

TALKLY keeps the conversation moving, with each person in their own language.

How it works

Four steps, at the pace of conversation.

Nothing to learn. People simply talk.

  1. 1

    Listens

    Picks up speech without interrupting the conversation.

  2. 2

    Understands

    Keeps the sentence, its intent and its context.

  3. 3

    Translates

    Moves into the other person’s language with a natural phrasing.

  4. 4

    Delivers

    Carries it to the other person in a natural speaking rhythm.

Speakers connect through headphones, a phone or a shared screen.

The product

One conversation, two languages.

Each person speaks and hears only their own language. Choose a scene and a language pair to see how a short exchange reads on each side.

Speaker ATürkçe

Says

Kapı nerede?

Hears

Düz gidin, ikinci koridordan sola dönün.

Translation

Speaker BEnglish

Hears

Where is the gate?

Says

Go straight, then turn left at the second corridor.

Illustrative example. Nothing is recorded or sent from this page.

Naturalness

It should sound like a conversation, not a device.

Translation is only useful if neither side has to slow down for it.

  • 01

    The conversation doesn’t break.

    Translation carries the meaning without waiting for the sentence to end.

  • 02

    The voice doesn’t sound read aloud.

    Emphasis and courtesy come through with the words.

  • 03

    Each side stays in their own language.

    Nobody has to switch to a shared language to be understood.

Conversation flow
  1. voice

    Spoken in Turkish

  2. meaning

    Intent and context kept

  3. translation

    Phrased in English

  4. voice

    Heard in English

Use cases

Every call center. Every front desk.

Wherever there is a call center, TALKLY belongs there. The same layer works face to face.

Call centers

Agents answer every caller in the caller’s own language, on every line, without a separate team for each language.

  • 01

    Airport

    Passengers get gate and transfer directions in their own language, the first time.

  • 02

    Hotel

    Check-in, requests and local advice happen at the desk instead of through gestures.

  • 03

    Hospital front desk

    Patients are directed correctly without waiting for an interpreter to become free.

  • 04

    Meetings

    International teams talk directly, without a pause for every sentence.

  • 05

    Retail welcome

    Visitors ask what they actually want to ask, and get a clear answer.

  • 06

    Tourism

    Guides and guests share the story, not just the route.

Languages

More than 90 languages, in both directions.

Every pair works both ways: either person can speak first.

90+languages

More than 90 languages are supported. The full list is shared during the demo.

TALKLY

Most requested

  • TürkçeTurkish
  • EnglishEnglish
  • العربيةArabic
  • DeutschGerman
  • FrançaisFrench
  • EspañolSpanish
  • РусскийRussian
  • 中文Chinese
  • 日本語Japanese
  • 한국어Korean

Everywhere

One conversation model, on any surface.

The way TALKLY works stays the same; only the screen changes.

  • 01

    Phone in hand

    For conversations on the move: a corridor, a gate, a taxi rank.

  • 02

    Tablet on the desk

    For a reception or counter where two people face each other.

  • 03

    Computer

    At a call-center desk with a headset: agent and caller each speak their own language.

Trust

Built for organizations that answer for their conversations.

Clear boundaries, decided by the organization that deploys it.

  • 01

    Purpose-bound

    Conversation content is used only to deliver the translation.

  • 02

    Your recording policy

    Whether anything is recorded follows the organization’s own rules.

  • 03

    Your environment

    In enterprise setups, models can run in the customer’s environment.

  • 04

    Role-based access

    Access is limited by role.

FAQ

Questions we hear often.

What happens if both people speak at the same time?

TALKLY is designed for turn-taking conversation, the way people naturally talk across a desk. When voices overlap, it follows the conversation rather than forcing a rigid order. How overlap is handled in your setting is clarified during the demo.

Which languages are available?

TALKLY supports more than 90 languages, including Turkish, English, Arabic, German, French, Spanish, Russian, Chinese, Japanese and Korean. Every pair works in both directions. The full list is shared during the demo.

Is the translation also shown as text?

TALKLY is voice-first, and text can be shown alongside the voice on a screen. Which surfaces show text in your setup is clarified during the demo.

Are conversations stored?

That is the organization’s decision. Recording follows your own policy, and conversation content is used only to deliver the translation. Retention details are clarified during the demo.

Can we run it on our own servers?

In enterprise setups, the models can run in the customer’s environment. Requirements depend on your scale and are clarified during the demo.

How do we get a demo?

Send the form below with a few words about where you’d use TALKLY. We will reply to arrange a time.

Demo

See TALKLY in a real conversation.

Tell us where two languages meet in your organization.

Optional. Languages, location, number of desks — anything that helps.

We use these details only to reply to your request.