Video summary
AI WhatsApp Balas Chat Otomatis 24 Jam
Main summary
Key takeaways
Product Being Reviewed
A “WhatsApp Automation” system for creating an AI WhatsApp chatbot/assistant that can:
- Auto-reply 24/7
- Manage multiple assistant accounts/numbers
- Escalate to human support when the AI can’t answer
The product is sold as source code.
Key Features Mentioned (Unique Points)
-
WhatsApp auto-reply / automation The bot replies to WhatsApp messages automatically using an AI assistant.
-
Multi-account / multi-number management
- Link multiple WhatsApp numbers
- Create multiple assistants (CS users)
- Only connected assistants/numbers go online
-
WhatsApp linking via QR code Connect each assistant/WhatsApp number by scanning a QR code (or using an installation code).
-
Assistant persona (“who is the AI?”) You define an assistant name/persona, and the bot answers questions accordingly (e.g., “What’s your name?”).
-
Configurable response style
- Language/style settings (e.g., predictive, formal/casual)
- Response length: short / balanced / detailed
-
Knowledge training (manual Q&A and website import)
- Add knowledge as question/answer pairs
- Paste content from a website and process it
- Select which knowledge items to practice/train
- Use recommendation mode so the bot prioritizes “most important knowledge” and skips less important items
-
Knowledge review & publish workflow Review trained knowledge before activation, then set it to Active for immediate use.
-
On/off switch for automatic AI answers “Automatic reply” can be turned off—when enabled, AI responses are active.
-
Real-time inbox integration A dedicated inbox in the dashboard shows WhatsApp message history in real time.
-
Escalation to human (CS takeover)
- If the AI can’t answer, the question is sent to a CS/human area
- Notifications help users avoid constant dashboard monitoring
- After escalation, the AI can stop answering further until handled
-
Follow-ups / reminders
- Create follow-ups targeting a number and specific content
- Use automatic sequence/flow to send follow-ups after pausing for a set number of hours/days
-
CRM integration / lead status tagging
- Linked numbers appear in CRM
- Leads are labeled hot/warm
- Warm = interested but not purchased
-
Message variation (“randomization”) Sends message variants to reduce spam-like repetition.
-
Targeting via contacts, labels, and groups Targets recipients from:
- WhatsApp contacts
- WhatsApp labels
- WhatsApp groups
-
CS rotator (number forwarding if blocked) If a blasting/assistant number is blocked, forwarding/continuation can use another connected number.
-
Blasting / sending controls
- Minimum interval between messages: mentioned as at least 40 seconds, with 60 seconds (1 minute) recommended
- Typical upper range: 70–80 seconds
- Pause after sending a certain number of messages (e.g., stop after every 20/30 messages)
-
Scheduled sending (“scheduled blast”)
- Schedule reminders/blasts for a specific day/time
- Supports attachments (picture/video/etc.)
- Emphasized that it can run long-term as long as the server/source code runs (24/7)
-
Scheduled Story/status uploads Automatically upload WhatsApp status at a chosen time/date.
-
Keyword-triggered flows (menu/buttons)
- Trigger word sends interactive buttons (e.g., “menu”)
- Buttons can drill down (e.g., “View services” → “Service hours”)
-
AI engine / model integration (DeepSeek mentioned)
- Choose an engine/model such as DeepSeek (described as cheaper/simpler), or alternatives (e.g., OpenRouter)
- Configure tokens/keys via platform top-ups
- Optional vision mode to read images (noted as “cheap” in subtitles)
-
External API / website integration (API integration) Documentation exists to send messages via link/code, including mention of OTP use for websites.
-
Group moderation rules
- Automatically remove/block spammy content or suspicious keywords/links (e.g., spam/gambling links)
- May require confirmation or take action depending on configuration
- Mentions deleting such messages/entries from dashboard/group management
Pros (Implied by the Video)
- 24/7 automation with reduced need for manual monitoring (notifications + server running)
- Escalation workflow: bot stops and hands off to humans when it can’t answer
- Highly configurable: persona, tone, length, knowledge sources, language style
- Lead management with CRM hot/warm tagging
- Reduced spam risk using randomized message variations and controlled sending intervals
- Scales across multiple assistants/numbers, including backup support via rotator
Cons / Limitations Mentioned or Implied
- Requires connecting/linking WhatsApp via QR code and managing connection status (some assistants may be offline until connected)
- Knowledge correctness matters—you must review/correct training knowledge to ensure intended answers
- Reliability depends on server uptime (automation runs only while the source code/server is running)
- AI may stop replying after escalation until CS/humans take over
- Scheduled blasts require timing discipline to avoid restrictions/spam-like behavior
User Experience / Setup Flow (From Subtitles)
- Purchase/install source code
- Login to the dashboard (admin credentials mentioned)
- Link WhatsApp (QR scan / installation code)
- Add assistant users (CS accounts) with WhatsApp number + access mode
- Configure assistant persona and answer style
- Train/publish knowledge (manual Q&A and/or website paste, review, activate)
- Turn AI auto-reply ON
- Use inbox + CS notifications for questions AI can’t answer
- Configure follow-ups, CRM, blasting rules, scheduling, and keyword flows as needed
- Integrate AI model engine (e.g., DeepSeek), optionally enabling vision mode
- Optionally integrate via API and set group moderation rules
Comparisons Made
- No direct comparisons to specific competing products/brands.
- Only an implied model choice comparison:
- DeepSeek is suggested as cheaper/simpler than OpenRouter.
Ratings / Numerical Scores
- No star rating or overall numerical review score is provided.
- Only quantitative guidance appears for messaging cadence:
- Minimum interval: at least 40 seconds / 60 seconds (1 minute)
- Typical max: 70–80 seconds
- Mentions pausing after ~20–30 messages
Overall Verdict / Recommendation
Recommended if: You want a self-hosted WhatsApp AI automation system (source code purchase) with:
- Multi-number support
- Knowledge training
- Scheduled campaigns
- CRM tagging
- Human handoff when AI can’t answer
Not ideal if: You prefer a simple plug-and-play hosted service, or you don’t want to handle setup/training—or manage uptime so the server/source code stays running.
Speakers / Multiple Views (If Applicable)
Only one main speaker is present in the subtitles; no distinct multi-speaker perspectives are clearly separated.