Vaani turns any moment of "I don't know anything here" — before a shift, before a meeting, after a lecture, in a new city — into clear words and a visual map. Spoken back in your own cloned voice. Free to start. Lives in your browser. Your data stays on your phone.
Every other tool makes you do the work — picking the language pack, choosing the model voice, writing the prompt, mapping the notes. Vaani does the work and hands you confidence. Five easys. One app.
30 seconds, voice-first, no signup. The app speaks the first instruction in your language. Reading is optional.
Hold the mic, say your name and favourite food. Your voice ready in 15 seconds. Opt-in, deletable any time.
One tap → your survival kit, your room kit, or your lecture kit. No menus to hunt.
Drop a podcast, a contract, or a list of confusing terms. Get plain language — contextual explanation only.
A one-page map of your terms — cause, effect, your scenario — tap any node, hear it back in your own voice.
Pick a scenario. Vaani tailors what it gives you to that exact moment. Voice + Scenario — the two personalization vectors no other tool combines.
The backend doesn't care if your input is a Portuguese taxi conversation or a metabolism lecture. Capture → Synthesize → Reflect. Output is always tailored to your voice and your scenario.
Audio, link, paste, or just pick a scenario. Mic, YouTube, podcast, contract, doctor's report, or a list of terms you keep mixing up. Held in your phone's RAM only.
Engine extracts the 15–20 core terms, maps how they connect, and translates the right vocab for who you are and where you stand. Indian pairs via Bhashini; rare pairs via Inworld.
An interactive visual map on screen + the summary line spoken back in your own cloned voice. Tap any node to see what triggers what. Save the map locally to your phone, or discard.
Pillar 1 is the moat — the recurring panic that pays the bills. Pillar 4 is the hook — the screenshot that spreads. Pillars 2 and 3 emerge as users discover the rest. Same engine. Same pipeline. Four faces.
The Mumbai driver before his shift. The Bengaluru nurse before her patient. The Gulf migrant on day one of a new site. 10 / 25 / 50 / 100-word survival packs tied to a specific corridor and a specific occupation — recurring need, recurring use.
The introvert founder before a VC pitch. The relocator before the izakaya dinner. The sales rep before the procurement meeting. Enter your role · their role · the setting → get 10 insider buzzwords, 5 hot news items, 3 bridge openers, 1 exit line.
The bus where you don't speak the language. The cafe where you're not sure if your order was called. Scrolling transcript on screen (read like you're scrolling social media) or contextual whispers to one earbud ("the barista just called your number"). Foreground-only in v1 — iOS Safari blocks background capture.
Drop a lecture, a podcast, a contract, or just a list of terms you keep mixing up. Get a one-page visual map of how 15–20 nodes connect, then tap "Hear in My Voice" — your own cloned voice speaks the summary. Drop new content next week → it auto-wires into your existing map.
Indian-language pairs route through Bhashini (Government of India, free, 22 scheduled languages) — best-in-class for Indic. World-language pairs and rare scripts route through OpenAI / Inworld. Engine choice is invisible to the user.
Pricing is corridor-aware — Mumbai drivers and Germany-bound nurses face different realities. PWA payments via UPI, mobile wallet, Stripe. App store fees bypassed by design.
Enough to feel the magic. Get hooked or walk away — no signup wall.
For the panicked shift day, the long meeting, the first day in a new city. Pay once, unlocked for 24h.
Self-voice cloning + cross-scenario knowledge graph that compounds on your phone over time.
One-off founder price. Locked at Monthly tier for life. First 100 only.
Vaani is a stateless app. Audio passes through. Your personal knowledge graph lives on your phone. Voice cloning is opt-in. Honest about what we don't control — third-party AI APIs have their own retention policies.
The connections you build over time — terms, nodes, relationships, your voice profile pointer — stored in your browser's local storage. Never uploaded to Vaani servers. Export to JSON any time. Wipe in one tap.
Audio is held in your phone's RAM only, sent to a transcription API, and discarded. Vaani operates no analytics, no behavioural tracking, no advertising tags. We never sell or share your data.
Bhashini (Govt of India), OpenAI, and Inworld process audio per their own published policies. OpenAI default retains inputs up to 30 days for abuse monitoring. Switching to enterprise zero-retention requires a higher tier we will move to as we grow.
Vaani takes any moment of "I don't know anything here" — before a shift, before a meeting, after a lecture — and returns three things tailored to you: the right words for your scenario, a visual map of how they connect, and the answer spoken back in your own cloned voice. One app instead of juggling Translate, Babbel, ChatGPT and NotebookLM. Voice-first. Mobile-first.
No. Vaani is a recurring-utility tool for the urgent moment — the 15 minutes before the shift, the meeting, the lecture, the doctor's visit. Language-learning apps lose roughly half their users by day seven because mastery is a marathon. Vaani solves the panic, not the marathon. You return because the panic returns.
Cognitive psychology calls this the production effect — hearing yourself say a complex word or phrase aids memory significantly more than hearing a synthetic voice. It also kills imposter syndrome. The first time a Mumbai driver hears himself say a Marathi phrase correctly, something shifts. That's the Vaani moment.
28 Indian languages — the 22 scheduled languages of India plus six widely-spoken non-scheduled (Bhojpuri, Awadhi, Magahi, Rajasthani, Tulu, Khasi). 10 world languages — English, Mandarin, Spanish, Arabic, Brazilian Portuguese, Russian, Japanese, German, French, Indonesian. Indian pairs run on Bhashini for free, government-grade quality. Premium voices route via Inworld.
Audio is processed in your phone's memory and the third-party API call, then discarded. Vaani servers store nothing. Your personal knowledge graph lives in your browser storage on your own phone, not on our servers. Voice cloning is opt-in, consent-gated, deletable any time. Third-party APIs (Bhashini, OpenAI, Inworld) handle audio per their own published retention policies — we are honest about what we cannot control.
Google Translate and Apple Live Translation translate words for you. Vaani tailors to two things they ignore — your scenario (the same word means different things at a Paris cafe vs a Mumbai meeting) and your voice (the cloned own-voice playback that locks new vocabulary into memory). All-in-one: vocab + voice + visual map + ambient understanding in one tap.
Free tier: 10 words a day, 1 scenario, basic map. Daily Pass: ~$0.25 / ₹20 via UPI or mobile wallet for 24-hour unlimited access. Monthly: ~$4.99 / ₹99 for unlimited use plus self-voice cloning. Founding-100 lifetime monthly for early supporters at a one-off founder price. Pricing differs by corridor — India tiers are micro-pay; Germany / Gulf corridors price for higher willingness-to-pay.
Progressive Web App, installable to home screen on any phone, no app store required. This keeps Vaani free of the 15–30% app-store cut and reachable via a single shared link. Some advanced features — continuous background ambient listening — are not possible inside a PWA on iOS Safari and ship foreground-only in v1.
Raj Singh, solo founder based in Sydney. Vaani is part of the Artigellence family alongside AAA (AI automation for Australian SMEs), synARKa Intelligence (geopolitical consensus engine), and Aarambh — The Founder Forge. Direct: raj@artigellence.com.
All policies below are current at June 2026, drafted in good faith, and subject to legal review and refinement. Vaani is in beta launch as part of the Artigellence family. For any policy query, email raj@artigellence.com.
Vaani is a cognitive aid and contextual explanation tool, not a licensed health, legal, financial, immigration, or professional advisor. Information generated by Vaani — including translations, visual maps, term explanations, and own-voice playback — is provided for educational and contextual understanding only.
Vaani provides no recommendations, no diagnoses, no certifications, and no advice. Health, Law, and Finance term-mapping is an explanation utility — it does not replace consultation with a qualified professional. Output may contain errors and should be cross-checked with primary sources before any consequential decision.
The Maharashtra Marathi driver mandate, Gulf Arabic compliance, Germany B1/B2 nursing requirement and similar regulatory hooks are marketing context, not legal guarantees. Vaani is a learning aid; it does not substitute for or guarantee passage of any official certification, test, or government-recognized proficiency examination.
By using Vaani you accept that this is beta-launch software, that AI output may be wrong, and that all consequential decisions remain your responsibility.
Who collects your data. Artigellence Augmentation Aggregator, Sydney NSW Australia. Direct contact: raj@artigellence.com.
What this site collects. Standard request headers and IP address sent by your browser to the hosting provider. If you join the waitlist or contact directly, the email and message you provide are stored.
What this site does NOT load. No Google Analytics, Plausible, Hotjar, Microsoft Clarity, Facebook Pixel, or LinkedIn Insight Tag. No behavioural tracking. No advertising network tags.
Your personal Vaani graph (knowledge graphs, voice-profile pointers, scenario history) is stored in your browser's local storage on your own device. Vaani does not upload, replicate, or back up this data to any server. If you clear browser storage or change device, your graph does not follow — it is yours alone.
Audio handling. Audio captured by the mic is held only in your phone's RAM during a session, transmitted to the chosen third-party transcription service (Bhashini, OpenAI, or Inworld), and discarded. Vaani servers retain no audio.
Your rights. Under Australia's Privacy Act 1988 and India's DPDP Act, you may request access, correction, or deletion of any personal data we hold (which, for the marketing site, is limited to your contact emails if you have written in).
Acceptance. By using vaani.artigellence.com, you accept these Terms.
Purpose. This site presents Vaani, a contextual cognitive-aid PWA. It is not a contract. Paid engagements are governed by the in-product purchase terms displayed at point of purchase.
No warranties. All content and AI-generated output is provided "as is" without warranty of accuracy, completeness, or fitness for any particular purpose. Do not rely on Vaani output for consequential decisions without independent verification.
Beta-release notice. Vaani is in beta. Pages, features, pricing, and policies are subject to change. The version current at the time of your purchase governs that purchase.
Jurisdiction. Governed by the laws of New South Wales, Australia. Disputes are subject to the exclusive jurisdiction of the courts of New South Wales.
What this site loads from third parties. Google Fonts (Fraunces, JetBrains Mono, Inter, Noto Sans Devanagari) serves typefaces. Google logs requesting IPs for font delivery but does not set advertising cookies for font requests, per Google's stated policy.
What this site does NOT load. No analytics. No advertising networks. No social-media embed cookies. No third-party trackers.
What Vaani stores in your browser. Your personal knowledge graph and voice-profile pointer are stored in browser local storage on your own device. This is not a cookie in the tracking sense — it is your data, on your phone, accessible only to you. Clear browser storage to wipe.
Voice cloning is opt-in. You explicitly choose to record a 15-second reference clip. The reference is sent to a third-party voice provider (Inworld or equivalent), which returns a voice profile token. Vaani stores only the token, not the reference audio.
Your voice profile is deletable. One tap from the privacy panel deletes the token from your device and revokes the voice profile with the provider. No residual on Vaani infrastructure because we never had any.
Biometric law awareness. Voice is biometric data under EU AI Act (effective Aug 2026), Illinois BIPA, and similar frameworks. Vaani operates an explicit consent flow and does not use voice data for any purpose beyond playback to you. Voice profiles are never shared, sold, aggregated, or used to train models.
Children. Vaani voice-cloning is not offered to users under 18 without verifiable adult consent.
Free tier. Nothing to cancel. Close the tab.
Daily Pass. Single-use 24-hour unlock, non-refundable once activated. Buy when you need it, use it.
Monthly subscription. Cancel any time from the in-product billing panel. Cancellation takes effect at the end of the current billing month — no recurring after that. Refunds for unused portions of a month considered case-by-case if the cancellation is within 7 days of charge.
Founding 100 lifetime. One-off purchase. Refundable up to 14 days from purchase if you have used less than ~10% of monthly equivalent features. Beyond 14 days, non-refundable but transferable once to another person you nominate.
How to claim. Email raj@artigellence.com with your purchase reference. Honoured within 7 business days.
Join the waitlist for Vaani at launch, or send me your scenario — the one moment of "I don't know anything here" that you wish had a button. Direct reply from Raj within 24–48 hours. No junior, no funnel sequence pretending to be a human.