Inbox
Every conversation auto-summarized when it ends. Agents skim a paragraph instead of reading 50 messages. CRMs get clean structured data. Handoffs become effortless.
On every convo
No setup needed
Per summary
Generated instantly
Time saved
Vs reading full transcript
Output
Goes to CRM cleanly
When a conversation ends (bot completes, agent resolves, contact closes), an LLM generates a 3-5 sentence summary + extracted fields. No agent action needed.
Summary contains: 1-paragraph narrative, customer intent, outcome, key extracted info (order ID, dates, amounts), recommended next action. Best of both formats.
When bot escalates, agent inherits the summary first. Read 30 seconds instead of scrolling through 8 messages. Faster context, faster resolution.
Summary auto-syncs to your CRM (Salesforce, HubSpot, Edesy CRM, or via webhook). CRM activity log gets structured WhatsApp data, not raw transcripts.
Search by summary content: 'show all conversations about refunds last month'. Faster than text search across full transcripts.
Summaries generated in your team's working language (English typically), even if the conversation was in Hindi/Spanish/Tamil. Automatic translation built in.
Customer chats with the bot or an agent. Could be 5 messages or 50. The bot uses tools, captures variables, possibly hands off to an agent.
Either the bot reaches an end state, the agent marks it resolved, or the conversation goes idle for the configured timeout (default 30 min).
An LLM reads the full conversation transcript + captured variables + any tool outputs. Generates a structured summary: paragraph + bullets + extracted fields + next action.
Visible in the conversation panel. Searchable. Exportable. Used in handoff briefings. Shown in contact's history timeline.
If CRM integration is configured, the summary + extracted fields push to the CRM as an activity record. Sales reps see WhatsApp activity alongside email and call notes.
| Feature | Edesy | Typical competitor |
|---|---|---|
| Auto-generated summaries | ||
| Structured + freeform format | ||
| Extracted fields (order ID, etc.) | ||
| Multi-language → English summary | ||
| Bot-to-human handoff context | Full transcript | |
| CRM auto-sync | Manual export | |
| Searchable across summaries | ||
| Cost per summary | ~₹0.05 | N/A |
Agent picks up a handoff conversation. Reads the 4-sentence summary in 20 seconds. Knows the customer's intent, what's been discussed, what tool calls succeeded/failed, what's next.
Handoff resolution time: 4 min → 90 sec average
Salesforce activity log shows WhatsApp summaries alongside emails and calls. Reps see entire customer history in one place — including non-Edesy conversations.
Sales rep reporting time saved: ~2 hrs/week per rep
Supervisors review agent conversations via summaries (faster than reading transcripts). Identify coaching opportunities at scale.
Coaching covers 100% of agents (vs ~20% via random sampling)
Export all summaries from a period for analysis. Common patterns emerge: top inquiry types, common complaints, repeat-customer themes.
Product team identified 3 friction points hidden in chat data
Returning customer? Agent sees the last 5 summaries in a glance. Knows the customer's history without reading 200 messages.
Customer recognition + personalization on every contact
For regulated industries: summaries serve as the auditable record. Includes flagged content (refund issued, claim approved, escalation reason).
Compliance review time reduced 80%
Free workspace, summaries on by default. Most teams notice the productivity lift within their first 50 conversations.