How an AI Assistant Replaced a Manager at Video Studio
Contents
A video studio in Bali was receiving booking requests around the clock — from Instagram, Telegram, and WhatsApp, in five languages. Managers couldn't keep up, and clients were leaving. We built a system that guides clients from the first message to a confirmed, paid booking — entirely within the messenger window.
The Problem: 3 AM Messages and Lost Clients
Villo Studio is a video production studio in Bali that works with tourists, bloggers, and brands from around the world. Clients message from different time zones — often late at night or early morning Bali time. Inquiries arrive in English, Russian, Indonesian, Chinese, and more. Managers simply couldn't respond fast enough.
The result: lost leads, missed bookings, negative reviews. The booking process itself was entirely manual — a manager would chat back and forth, suggest time slots, send payment details, wait for a receipt screenshot, verify it visually, and log everything in a spreadsheet. Errors piled up.
Client messages in → waits until morning → leaves for a competitor. No unified system: conversations in one place, payments in another, client records in a third.
The Solution: 5 Modules, One Platform
We didn't plug in an off-the-shelf chatbot — we designed a comprehensive system from scratch, combining an AI assistant, payment logic, CRM, and even live camera monitoring in a single dashboard.
1. Multilingual AI Assistant
The assistant is connected to Instagram, Telegram, and WhatsApp simultaneously. It auto-detects the client's language and communicates fluently in English, Russian, Indonesian, Chinese, and more. It handles initial consultations, describes services and pricing, and shows available shooting slots. It operates 24/7 — even at 3 AM, the client gets a full, contextual response within seconds.

This isn't a "dumb button bot." It understands context, handles complex questions, and can even negotiate special conditions. When a situation requires a human, an operator can jump in with one click.
2. Automated Booking & Payment
Right inside the chat, the assistant offers available slots from the calendar, processes the booking, and automatically sends a payment link connected to the payment gateway. After payment, the client sends a receipt screenshot, and the system verifies it using computer vision: it recognizes the amount, date, transaction status, and automatically confirms the booking.
The entire journey — from first message to confirmed, paid reservation — happens in one messenger window. No website redirects, no forms, no waiting for manual confirmation.

Our client, a video studio in Bali, kindly allowed us to test the service. Go ahead and try contacting them in any messenger, and our assistant will respond.
3. Dialog Management Panel
A proprietary platform — Tantal AI Platform — where operators see all assistant-client conversations in real time. At any moment, a human can take over a chat, handle it manually, then hand it back to the bot with a single button. Three modes: Bot, Operator, Manual.
This is critical for service businesses — the bot handles 80-90% of standard questions, but when a client wants to discuss a custom project, the manager steps in instantly with the full conversation history in front of them. The client doesn't even notice the handoff.
4. Internal CRM System
Every client who messages through any channel automatically enters the CRM with their own profile card and sales funnel. Cards are populated automatically:
Pulled from the messenger profile
Auto-detected from the first message
Where they came from: ads, referral, organic
Last shoot date, total spend, booking history
This allows the marketing team to segment their audience and run precisely targeted campaigns: retarget those who didn't complete payment, offer a follow-up shoot two weeks later, provide special conditions for high-value clients.
5. Live Studio Camera Feeds
Live feeds from studio cameras are integrated directly into the CRM panel via RTSP → HLS streaming. The owner can monitor studio load in real time, oversee ongoing shoots, and respond to situations instantly — even from another city.
Tech Stack
React + TypeScript
Node.js, Python (FastAPI)
Custom OCR model for payment receipts
Instagram API, Telegram Bot API, WhatsApp Business API
PostgreSQL, Redis
RTSP → HLS, WebSocket
What This Means for Your Business: A Practical Breakdown
You can build an AI assistant like this for your business if you face similar challenges. Here's what to pay attention to:
Messengers as a unified sales channel. Clients don't want to visit a website, fill out forms, or wait for a callback. When the entire journey — from consultation to payment — happens in one messenger window, conversion rates multiply. In Villo Studio's case, clients complete bookings without ever leaving the chat.
Computer vision for verification. Automated receipt verification through OCR isn't rocket science — it's a very achievable solution. The system recognizes the amount, date, and status, then confirms the booking without human involvement. This eliminates delays and manual errors.
Hybrid model: bot + human. The key isn't replacing the manager entirely — it's giving them a tool. The bot handles 80-90% of standard inquiries, and for non-standard cases, the operator joins instantly with the full conversation history. The client doesn't even notice the transition between bot and human.
Auto-populated CRM as a marketing foundation. When client cards fill themselves automatically (source, language, history, spend), marketing campaigns become dramatically more effective. You know exactly who to re-engage, who to push through email, and which channel brings your most valuable clients.
Results
The studio now accepts inquiries around the clock without adding headcount. Clients complete the full journey from first message to paid booking within the messenger, never leaving the conversation. Booking conversion increased, lead loss dropped significantly, and marketing campaigns became more precise thanks to automatic data collection from every client interaction.
Read also
How an AI Assistant Handled 10,000+ Tickets Per Day
KupiCode was processing thousands of customer requests daily — payment issues, order statuses, refunds. The support queue kept growing, and average response time…
Automating Dental Patient Scheduling with AI Assistant
A dental clinic administrator spends up to 40% of their working day on phone calls — and one in five patients still doesn't show up. Money is spent, the chair sits…
AI Voice Robot for Business: Automation in Bitrix24 and 1C
Tantal: AI-Powered Voice Robots for B2B Your manager spends a third of their working day on calls, with scripts that haven't changed in two years. A voice robot can do…
AI Voice Bot for Auto Shop: Stop Missing Calls
Up to 40% of inbound calls to auto repair shops come in outside business hours — according to Calltouch data. Most of them go unanswered. You have already paid for…
Voice Robot for Telemarketing Calls: How It Pays Off in Small Business
Your manager spends half a day on calls that the robot makes at night. Do you know how much each of these calls costs? Most likely, there's no exact figure. And it…
AI Development for Business: How Not to Overpay and Get Results
Business owners love the statistic "97% of companies are already implementing AI." But there's a second one that's less often mentioned: only 36% can prove that AI has…