Artificial Intelligence

AI reporting, channel assistants, chatbot flows, conversation summaries and HR screening, on one AI credit balance.

The MyVoipCRM artificial intelligence stack runs on a single AI credit wallet across the whole CRM. Generating reports in natural language, answering automatically on channels like WhatsApp and Web Chat, designing chatbot flows, giving agents a conversation summary or scoring job candidates in HR — all live in the same ecosystem. Each feature is configured from its own menu; consumption is tracked on the AI Credits screen.

What it does

  • AI Report: customer, company, call-record, sales, support, conversation and quote reports from a Turkish/English natural-language request; PDF download, table and chart options
  • Data chat: instant answers to short data questions such as “How many customers last month?” (separate from full-report mode)
  • AI assistant (per channel): knowledge-base automatic replies for WhatsApp, Instagram, Facebook Messenger and Web Chat; website crawling, document upload, a system prompt
  • Take from AI / hand to AI: control passing between the agent and the AI inside customer conversations
  • Classic AI chatbot: intent recognition, working-hours mode, session timeout and a system prompt inside the channel integrations
  • Chatbot flows: message, question, condition, form and tag steps in a visual flow editor; WhatsApp, the Meta channels and Web Chat
  • Flow analytics: starts, completions, drop-off points and node usage statistics
  • Conversation summary: an AI summary in WhatsApp, Instagram, Facebook, Web Chat, SMS, email and internal chats
  • HR candidate evaluation: AI scoring of job-posting candidates against the position; single and bulk evaluation
  • AI credits: balance, last-30-days consumption, per-feature breakdown, credit purchase, low-balance notifications
  • Usage and cost: token consumption and estimated cost (USD) analysis in the channel AI assistants

Menu and screens

  • AI Report — AI Report → AI Report (/ai-report)
  • AI credit balance — Accounting → AI Credits (/ai-kredileri)
  • Buying credits — Accounting → Buy Credits (/ai-kredileri/satin-al)
  • Chatbot flows — Conversations → Settings → Chatbot (/chatbot-flow)
  • Flow analytics — Conversations → Settings → Flow Analytics (/flow-analytics)
  • WhatsApp AI assistant — Conversations → Settings → WhatsApp → Integration → AI Assistant (/whatsapp-integration)
  • Instagram AI assistant — Conversations → Settings → Instagram → Messenger Integration → AI Assistant (/instagram-integration)
  • Facebook AI assistant — Conversations → Settings → Facebook → Messenger Integration → AI Assistant (/facebook-messenger-account/messenger-integration)
  • Web Chat AI assistant — Conversations → Settings → Web Chat → edit site → AI Assistant (/web-sohbet-siteleri)
  • Conversation summary — the in-chat “AI Summary” button on the channel tabs (/mesajlasma)
  • HR candidate AI evaluation — Human Resources → candidate record → AI evaluation tab

AI Report

With the AI Report module you request reports through a chat interface. The system analyses the request and, when needed, asks clarifying questions about the date range, breakdown and output format. There are two modes: Full report (HTML/PDF with table and chart options) and Data chat (fast numbers and summaries).

  • Quick templates: customers, companies, call records, sales
  • Supported data sources: call records (CDR), sales, agent tasks, support tickets, conversations, quotes, customers
  • PDF download; “continue” on a report for further analysis
  • Report history is kept locally in the browser (no permanent server archive)

The channel AI assistant

The AI Assistant tab in the WhatsApp, Instagram, Facebook Messenger and Web Chat integrations offers advanced configuration:

  • Seeding initial knowledge by crawling the business website
  • Contact details, product/FAQ specifics, a system prompt (behaviour rules)
  • Knowledge base: PDF and text documents (persistent RAG depth)
  • Priority mode: AI first, or chatbot flow only
  • Working hours, session duration and intent-recognition settings
  • Token counts and estimated USD cost on the usage-and-cost tab

In the Conversations screen an “AI replying” badge shows on customer conversations; the agent manages control with take from AI or hand to AI.

Chatbot flows and analytics

  • Design visual flows in the Chatbot Flow editor by channel type (WhatsApp, Facebook, Instagram, Web Chat)
  • Node types: message, question, condition, form, tag; drag-and-drop connections
  • AI-assisted flow routing and intent recognition can step into chatbot stages
  • Flow Analytics shows start counts, completion rates, drop-off points, form completions and tag statistics

The conversation summary

On every messaging channel (WhatsApp, Instagram, Facebook, Web Chat, SMS, email, internal chat) the AI Summary button in the chat header condenses the conversation into structured sections: the main topic, intent, requests, commitments, open items and next steps. Built for agents to grasp a long history quickly; it consumes AI credits.

HR candidate evaluation

For companies with the AI Report module active, an “AI evaluation” tab opens on Human Resources candidate records. Position-specific analysis for a single candidate or all candidates on a posting: a fit score, a suggested star rating, strengths/weaknesses and a summary. Bulk evaluation results are reported in a modal; the run stops if credits are insufficient.

AI credits

  • Current balance, total used and total purchased credits
  • Last 30 days: request count, tokens, credits spent, estimated cost (USD)
  • Per-feature consumption: channel replies, site crawling, intent recognition, conversation summaries, AI reports, embeddings, HR evaluation and more
  • Credit-package purchase (3D Secure payment)
  • A notification email for low or exhausted balance

Getting started

1. Credits and modules

  • Check your balance under Accounting → AI Credits; buy credits if needed.
  • The module for each feature you plan to use must be active (e.g. AI Report, WhatsApp, Web Chat).

2. AI Report

  • Open AI Report → AI Report (/ai-report).
  • Type your request or pick a quick template; choose Full report or Data chat mode.
  • Answer the clarifying questions; view the report or download the PDF.

3. The channel AI assistant

  • Open the AI Assistant tab on the relevant channel integration screen.
  • Follow the guided order: API key → website crawl → contact/details → system prompt → knowledge-base documents.
  • Activate the assistant; test it in Conversations and watch consumption under “usage and cost”.

4. Chatbot

  • Create a new flow under Conversations → Settings → Chatbot; pick the channel type.
  • Attach the flow to the channel from the integration or the Web Chat site settings.
  • Measure performance with Flow Analytics.

AI assistant vs. chatbot flow vs. classic AI chatbot. The AI assistant produces free-text replies from a knowledge base and a language model (the advanced training screen). Chatbot flows are pre-designed step-by-step scenarios. The “AI chatbot” on the integrations’ Chatbot & Working Hours tab is simpler intent-based automation. The three can run together or separately.

Who it is for

  • Managers who want reports without SQL or a report editor
  • Support teams answering customers 24/7 with the WhatsApp/Web Chat AI assistant
  • SMEs building out-of-hours automation on the chatbot + AI hybrid model
  • Call centres whose agents need long conversation histories summarised fast
  • HR teams speeding up candidate pre-screening

Frequently asked questions

An AI feature stopped / “insufficient credits” error

Check your balance on the AI Credits screen. When credits run out, conversation summaries, AI assistant replies and AI reports can stop. Top up by purchasing a package.

The AI Report menu is missing

The AI Report module must be active for your company and your account needs the AI Report permission.

The AI assistant keeps saying “visit our website”

Complete the assistant training checklist: contact details, specifics, the system prompt and at least one knowledge-base document. Automatic crawling can be limited on modern (SPA) sites — enter the critical information by hand.

The chatbot flow menu is missing

At least one messaging module (WhatsApp, Facebook, Instagram or Web Chat) must be active.

HR AI evaluation is unavailable

It requires the AI Report module and sufficient AI credits. The candidate must first be registered on a posting.

Which modules consume AI credits?

WhatsApp, Instagram, Facebook, Web Chat, internal chat (conversation summaries) and AI Report (reports + HR evaluation) sit on the AI credit ecosystem. Each feature’s own module prerequisite still applies.