← Back to talksPre-seed pitch · 2026

Liltra

The first AI coach that truly hears your voice — for actors, speakers, and anyone whose voice is their career.

·Present →Download slides (PDF)

Pre-seed · 2026

liltra

The first AI coach that truly hears your voice.

For actors, speakers, executives, and anyone whose voice is their career.

Arielle Ostankova · actress & engineer · arielle@liltra.com

Slide 1 / 11

Warm hand-off. Eye contact.

Hi — I'm Arielle. I'm building Liltra: the first AI coach that truly hears your voice.

For actors, speakers, executives — and anyone whose voice is their career.

I want to take you through why this is suddenly possible, who pays first, and where this goes.

The script wanted General American. I speak with a slight Eastern European accent. So I built my own loop.

Recorded takes, sent them to friends and Gemini, found the problem words, drilled them until my accent and delivery were on point.

It was inconvenient, slow. But it got me the role.

Liltra is that loop — streamlined and scaled.

Almost every actor, public speaker, executive, and interview candidate runs a similar hacky routine — because the tool that should exist doesn't. Yet.

Arielle Ostankova · Founder · Trained actor · 20 yrs software · Multilingual

Slide 2 / 11 · THE MOMENT I BUILT THIS

Slow. Personal. This is the founder story.

Last year I was prepping for an audition. The script wanted General American. I speak with a slight Eastern European accent.

So I built my own loop. Recorded takes. Sent them to friends. Asked Gemini what I was getting wrong. Drilled the problem words until my accent and delivery were on point.

It was inconvenient and slow. But it got me the role.

Liltra is that loop — streamlined and scaled.

Almost every actor, public speaker, executive, and interview candidate runs a similar hacky routine. Because the tool that should exist doesn't. Yet.

Three tiers of pronunciation pain.

Ranked by career impact of accent and pronunciation.

01 · Loud pain

Actors, YouTubers & influencers

Already paying $150/hr for coaches. One bad take — a lost role, a lost audience.

500K actors globally

02 · Known pain

Public speakers & executives

Before the investor call or the keynote, they rehearse the content. Nobody helps them rehearse the delivery.

30M professionals

03 · Silent pain

Interview candidates & mobile professionals

Rejected but never told why. Don't know the problem is fixable.

1.5B English learners

Start with actors — they know the pain the most and are ready to pay. Expand from there.
Slide 3 / 11 · THE INSIGHT

Map the three tiers of pain.

There are three tiers of pronunciation pain — ranked by career impact.

Tier 1 — actors, YouTubers, influencers. The loud pain. Already paying 150 dollars an hour for coaches. One bad take is a lost role, a lost audience. 500,000 actors globally.

Tier 2 — public speakers and executives. The known pain. Before the investor call or the keynote, they rehearse the content. Nobody helps them rehearse the delivery. 30 million professionals.

Tier 3 — interview candidates and mobile professionals. The silent pain. Rejected and never told why. Don't know the problem is fixable. 1.5 billion English learners worldwide.

We start with actors. They know the pain the most. They're already paying. From there we expand.

Upload your script. Get phoneme-level coaching.

1

Upload the script

Audition scene, TED talk, pitch deck, interview prep — any text you need to nail.

2

Record your take

Professional-grade analysis. We find the exact sounds working against you.

3

Targeted drills

Personalized exercises for your specific problem phonemes. Not generic — yours.

4

Measurable before / after

We quantify the delta. See improvement in real-time. That's the whole point.

Audition · Script #1

“I don't know whether I'm coming or going, but I know one thing — I'm not leaving this room.”

General American

Problem phonemes
Accuracy score72

/ð/ whether · /ɪŋ/ going

Slide 4 / 11 · THE PRODUCT

Walk through the four steps. Make it concrete.

Hear, diagnose, train. Upload your script. Get phoneme-level coaching.

Step one — upload the script. Audition scene, TED talk, pitch deck, interview prep. Any text you need to nail.

Step two — record your take. Professional-grade analysis. We find the exact sounds working against you.

Step three — targeted drills. Personalized exercises for your specific problem phonemes. Not generic — yours.

Step four — measurable before and after. We quantify the delta. See improvement in real time. That's the whole point.

Not a waitlist. A launch week.

April 2026

17 / 21 actors loved it

Out of 21 invited professionals, 17 said they would pay.

Live

Script vertical — shipped

Actors ran real audition scripts. Not a mockup — a working product they tested with their own material.

EN · DE

Two languages day one

English and German. Built for multilingual professionals from the start.

I uploaded my audition script. It figured out exactly where I'm tripping up — the sounds I didn't even know I was getting wrong. I'd pay for this.

Petro Stashenko · working actor

The friend I send my recordings to — this replaces that. Faster, more precise, and available at 2am before an audition.

TikTok creator · 50K followers

Slide 5 / 11 · TRACTION

Confident. The numbers earn it.

Not a waitlist. A launch week.

April 2026, prototype launch. Out of 21 invited professionals, 17 loved it and said they would pay.

Script vertical — shipped. Actors ran real audition scripts through it. Not a mockup. A working product they tested with their own material.

Two languages from day one. English and German. Built for multilingual professionals from the start.

Quote — I uploaded my audition script. It figured out exactly where I'm tripping up. The sounds I didn't even know I was getting wrong. I'd pay for this. — Petro Stashenko, working actor.

Quote — The friend I send my recordings to — this replaces that. Faster, more precise, and available at 2am before an audition. — TikTok creator, 50K followers.

The lab just became possible.

01 · 2025 Audio AI

AI can measure how you speak — not just what.

Until 2024, speech AI told you what you said. Now it measures how — rhythm, melody, accent, phoneme by phoneme.

02 · 2025 Voice AI

AI can show you the target — in your own voice.

Generate a reference take in any dialect, in a voice that sounds like yours. Hear what ‘correct’ sounds like before you practice.

03 · Empty today

Every tool targets learners. The pro lane is wide open.

Duolingo, ELSA, BoldVoice — all built for language learners. Actors, speakers, candidates: zero professional tools.

AI finally has ears. We built the pronunciation lab around them.
Slide 6 / 11 · WHY NOW

Three reasons. Three breaths.

The lab just became possible.

One — AI can now measure how you speak. Not just what you say. Until 2024, speech AI told you what you said. Now it measures how — rhythm, melody, accent, phoneme by phoneme.

Two — AI can show you the target in your own voice. Generate a reference take in any dialect, in a voice that sounds like yours. Actors can hear what correct sounds like before they practice.

Three — every tool targets learners. The professional lane is wide open. Duolingo, ELSA, BoldVoice — all built for language learners. Actors, speakers, candidates — zero professional tools.

AI finally has ears. We built the pronunciation lab around them.

No one built for the professional.

DuolingoELSAPrivate coachLiltra
Built for professionals~
Phoneme-level coaching
Script / presentation workflow
Measurable drill progress~
Available 24/7
Under $50/mo
$34/mo — custom voice pipeline, phoneme coaching, professional script workflow.
Slide 7 / 11 · COMPETITION

Read across the row. We win every line that matters.

No one built this for the professional.

Duolingo and ELSA are for learners. They don't think about scripts, presentations, or measurable drill progress.

Private coaches are great — but $150 an hour, by appointment, in one timezone.

Liltra — built for professionals. Phoneme-level. Script and presentation workflow. Measurable progress. 24/7. Under 50 dollars a month.

Thirty-four dollars a month. Custom voice pipeline, phoneme coaching, professional script workflow.

Solo founder. Full stack. By design.

Arielle Diana Ostankov · Founder & CEO

Founder–market fit

Professional actress, native EN/DE/RU, built her own pronunciation loop years before Liltra. 17/21 actors confirmed the problem in person.

Technical depth

20+ years full-stack engineering, M.Sc. CS, hands-on AI / NLP. Building the entire MVP solo — from phonetic engine to frontend.

Founder experience

Built startups before: Building Radar, All.Pics, Artware.Studio. Solo at this stage is by design — one person moving fast across product, engineering, and domain.

Ready to scale

Advisers already in place. Hiring begins immediately after close.

Slide 8 / 11 · WHY ME

Earn the trust line by line.

Solo founder. Full stack. By design.

Founder-market fit — professional actress, native English, German, Russian. I built my own pronunciation loop years before Liltra. 17 of 21 actors confirmed the problem in person.

Technical depth — 20+ years full-stack engineering. M.Sc. in computer science. Hands-on AI and NLP. Building the entire MVP solo — from phonetic engine to frontend.

Founder experience — built startups before. Building Radar, All.Pics, Artware.Studio.

Solo at this stage is by design. One person, moving fast across product, engineering, and domain.

Advisers already in place. Hiring begins immediately after close.

Dialect coaches charge $120–350/hr.

We charge $34/mo.

Free

$0

Weekly practice sessions — habit builder, top of funnel.

Premium

$34 / mo · $340 / yr

Full pronunciation lab. AI coaching plus phoneme-level feedback.

Enterprise

Custom

Teams, schools, agencies. Volume licensing.

Free users share progress clips — the product markets itself.
Slide 9 / 11 · BUSINESS MODEL

Anchor on the price gap.

Dialect coaches charge $120 to $350 an hour. We charge $34 a month.

Free tier — weekly practice sessions. Habit builder, top of funnel.

Premium — $34 a month or $340 a year. Full pronunciation lab. AI coaching plus phoneme-level feedback.

Enterprise — teams, schools, agencies. Volume licensing.

Growth flywheel — free users share progress clips. The product markets itself.

€500K

Turn 17 actors who love it into 500 who pay for it.

No per-session human cost. Every user served by the same pipeline. Margins improve with scale.

M4

Payments live, first 50 paying actors.

M8

200 paying actors, retention data in hand.

M12

500 paying actors = $17K MRR.

M15

Series A ready, 6-month retention proven.

500 actors at $34/mo with 85% margins. The only question left is how fast we reach 10,000.
Slide 10 / 11 · THE ASK

Land it cleanly. This is the number.

€500,000.

Turn 17 actors who love it into 500 who pay for it.

No per-session human cost. Every user served by the same pipeline. Margins improve with scale.

18-month milestones — month 4, payments live, first 50 paying actors. Month 8, 200 paying actors, retention data in hand. Month 12, 500 paying actors equals $17K MRR. Month 15, Series A ready, six-month retention proven.

Proven acquisition channel with CAC under $50. 500 actors at 34 dollars a month with 85% margins.

The only question left is how fast we reach 10,000.

A pronunciation lab for professionals.

The window just opened.

Actors first — they know the pain and already pay $120–350/hr for coaches.

Then speakers, creators, and the 1.5 billion learning a language.

Let's build the lab that scales.

Find your lilt.arielle@liltra.comliltra.com
Slide 11 / 11 · CLOSER

Soft landing. Eye contact.

A pronunciation lab for professionals. The window just opened.

Actors first — they know the pain and already pay $120 to $350 an hour for coaches.

Then speakers, creators, and the 1.5 billion learning a language.

Let's build the lab that scales.

Find your lilt. arielle@liltra.com. liltra.com.

Thank you.

Investor or partner?

Reach me at arielle@liltra.com or via linktree.