Better sales by your staff in real time.
ONVY is an AI audio badge for frontline staff in hospitality and retail. It listens, finds the slips, and turns them into better sales.
Don't name the price yet. Lead with the three main benefits.
- Championship
- Grand Prix Winner, HSE KLUCH 2025
- national superfinal, business track
- Pilots
- 12 restaurants
- incl. the Chaikhona №1 chain
- Raised
- $14.5K pre-seed
- grant and angel
- Signed
- LOI with RS Capital
- private equity fund
Pilots
- Chaikhona №1
- Miller
- Vasilki
- Grusha
- PIMS
- Khachapuri i Vino
- Tekhnikum
- Vino i Khinkali
Partners
One shift with ONVY
A new hire takes 5–8 weeks to sell at plan and loses 8–15% of deals in the conversation. Managers can't hear why. ONVY can.
- Guest
This one looks good. How much is it?
- Server
It's $24.
- Trigger
Price named before value. Badge vibrates.
- Hint
Don't name the price yet. Lead with the three main benefits.
Illustrative dialogue. The hint is the actual wording from the ONVY prototype, translated.
- Hears
- Records only in the sales area, streams speech to text.
- Coaches
- Checks each line against the script and 100+ triggers. On a miss: buzz and a better line.
- Trains
- After the shift, a lesson built from the employee's own conversations.
Offer dessert with coffee. Yesterday 6 of 9 guests didn't hear the offer. Two lines that work at table 6: listen, then try.
- 1Elizaveta P.112%
- 2Anastasia I.104%
- 3Alexey K.97%
Privacy by design
Built to Russia's personal data law: signs tell customers they're recorded, no voice biometrics, no recording in private areas, encrypted audio on servers in Russia.
Managers see coaching trends, not raw audio. Staff see only their own conversations.
What we don't know yet
Does a live hint teach the skill, or does performance drop when the badge comes off? Our next trial measures it.
Hardware and AI pipeline
- Board
- ESP32-WROOM-32
- Mic
- ST MP34DT05 MEMS
- Battery
- 500 mAh Li-Po, 12 h
- Link
- Bluetooth / Wi-Fi, microSD buffer
- Speech
- Yandex SpeechKit, streaming
- Script check
- ruBERT, real-time triggers
- Coaching
- YandexGPT hints and lessons
- Platform
- React web app, Yandex Cloud
Traction
Raised to date
$14.5K
$5.5K grant + $9K angel · LOI with RS Capital, a PE fund with retail chains in its portfolio
- 12
- restaurant pilots, from 3 venues to the Chaikhona №1 chain
- 33
- venues reached
- 40%
- of in-person visits agreed to a pilot, vs 7% of cold calls
- 330
- badge-minutes in the first field test, 11 triggers caught
What the pilots changed
- Microphones picked up background music
- Added software noise reduction
- Data dropped with the internet
- Added an SD-card buffer on the badge
- A flat fee per venue ignored seasonality
- Switched to pay-per-minute of active speech
- Managers had no time for analytics
- Personal training became the core product
Timeline
20 customer interviews at 9 chains, first financial model, team of four.
First field test in 3 venues: 14 conversations, 11 triggers caught, 330 badge-minutes. Grand Prix Winner at the national superfinal.
Investor video conference and follow-up calls. LOI signed with RS Capital.
Pricing rebuilt around minutes of active speech.
Pilot at Chaikhona №1, a Moscow restaurant chain. 12 pilots in total.
First production batch of badges and CRM connectors.
Superfinal awards, HSE University, Moscow, December 8, 2025.
Grand Prix of Russia's largest school case championship
- Grand Prix Winner, HSE KLUCH 2025 national superfinal#3 of 3,258 teams, business track
- Finalist, All-Russian Olympiad in Entrepreneurship2026
- Semifinalist, Startech PrizeBest technologies by students and school students, 2026
- Innovators Academy accelerator residentTop 100 startups in Russia, Bunch
- Featured by HSE UniversityStory about the team and ONVY, September 2026
“The business model the team proposes could be of interest to the B2B segment, in particular to companies like Rostelecom that are actively developing new lines of business.”
The team behind ONVY
Maksimilian Albeshchenko
Founder & CEO · Product, finance, fundraising
Why does a new server need two months to sell well, when the fix is one sentence at the right moment? I asked 20 baristas and managers, recruited the team, built the financial model and captained us to the Grand Prix win. When my teammates left for university, I kept going. The first US market I want to test is campus dining.
Achievements
- #1, IEO Business Case
- Top 50, Harvard Crimson Business Competition
- Absolute winner, All-Russian Olympiad in Social Studies
Experience
- Head of Business Development, Steel Robotics, a US robotics startup
- VC analyst intern, FRII Invest
2025 championship team, now advisors
-
Vladislav Fokin
CTO, 2025
Hardware and the AI pipeline. Business analyst at Strauss; winner, DEADLINE case championship.
-
Anastasia Naumkina
UX/UI & research, 2025
Customer interviews, field tests, product design. Intern at MOEX Group; winner, DEADLINE.
-
Dmitry Bronnikov
CBDO, 2025
Venue outreach and partnerships. Investment fund intern; finalist, Harvard Crimson Business Competition.
Pre-seed round
Raising $37.5K for 10% at a $337.5K pre-money valuation ($375K post) to ship the first production batch of badges and CRM connectors. Looking for an angel with restaurant or retail experience.
- $0.09
- per minute of analyzed speech; the badge sells at $38 at a $30 cost
- 80%
- gross margin on analyzed minutes; cost to serve ~$6 per employee per month
- $640 M
- addressable market in Russia, bottom-up from 2.41 M frontline employees
- Q2 2027
- break-even in the financial model; pre-revenue today
Use of funds
- Payroll
- $16K
- Marketing
- $10K
- Badge batch
- $6K
- Admin
- $5.5K
Badge batch: 200 badges × $30. Ruble figures converted at 80 ₽ per dollar.
Roadmap
First production batch; connectors for Bitrix24, amoCRM and 1C
Sber Z and MIPT accelerators, Skolkovo residency, patent for the badge
US pilot: campus dining and college-town restaurants, built with a student team on campus
Talk to us
A pilot takes one week: we bring the badges, you pick the venue, you get a per-employee report. Investors get the deck and the financial model on request.
Public sources
HSE University publishes results in Russian. Our team is listed there as «Чистая прибыль, приди» (team 1279), the product as ONVY.