API precision meets dashboard visibility
8 real workflows where developers and business users collaborate on the same data. Each scenario is grounded in a system engineering principle and shows exactly what each surface contributes.
CTWA Ad Lead Capture + Nurture Pipeline
A Facebook or Instagram Click-to-WhatsApp ad opens a conversation. The API captures the ad referral metadata and creates the contact. The dashboard agent sees the lead in the inbox and monitors sequence progress. When the lead needs human touch, the agent intervenes with full context.
API side
- Webhook receives inbound message with ad referral (source, headline, ad_id)
- POST /api/v1/contacts creates the lead with ad-source tags
- POST /api/v1/sequences/{id}/enroll starts the nurture drip
- Milestone events track engagement (replied, clicked, qualified)
Dashboard side
- Lead appears in Inbox under "Interactive" category (free 24h window)
- Agent reviews sequence progress and step status
- Manual reply when the lead asks a question mid-sequence
- Contact manager shows full history and tags from ad attribution
AI-First Support with Human Escalation
Your LLM handles first-line support via the API. When it cannot resolve, it assigns the conversation to a human agent. The agent picks up in the dashboard with full AI conversation history and uses built-in AI reply drafting for faster responses.
API side
- Webhook delivers every inbound message to your LLM endpoint
- LLM processes and replies via POST /api/v1/messages
- Confidence below threshold triggers POST /conversations/{phone}/assign
- GET /api/v1/conversations/{phone}/messages provides full context to the agent
Dashboard side
- Escalated conversations appear in the Inbox with "assigned to you" highlight
- Agent reads the full AI conversation thread for context
- AI reply drafting suggests responses grounded in conversation history
- Agent resolves, tags, and adds internal notes for future reference
Abandoned Cart Recovery with Live Agent Close
Your e-commerce platform detects an abandoned cart and triggers a WhatsApp reminder via the API, with consent verified first. When the customer replies with questions, a dashboard agent handles the conversation with product library context and closes the sale.
API side
- E-commerce webhook fires on cart abandonment
- GET /api/v1/contacts/{phone}/consent verifies marketing permission
- POST /api/v1/messages sends cart reminder with product image and checkout link
- Webhook captures the customer's reply
Dashboard side
- Customer reply surfaces in Inbox under "Interactive" (free reply window)
- Agent sees the cart reminder that was sent and product details
- Product library provides quick access to catalog for upsell
- Agent sends payment link and tracks conversion in cost dashboard
Programmatic Campaigns with Manual Override
Your backend creates segments and launches campaigns based on business logic (time-of-day, purchase history, inventory levels). The marketing team monitors performance from the dashboard and can pause underperforming sends, adjust audience, or manually intervene.
API side
- POST /api/v1/segments creates dynamic audience segments
- POST /api/v1/campaigns builds the campaign with template and audience
- POST /api/v1/campaigns/{id}/launch starts delivery
- GET /api/v1/campaigns/{id}/stats polls delivery metrics
Dashboard side
- The API-created send appears in the Broadcasts subtab with a status pill (sent, scheduled, paused)
- Marketing team pauses a send that's underperforming
- Broadcast history shows delivery counts for every send, API-triggered or not
- Reply conversations routed to inbox for manual follow-up
Both surfaces share the same D1 database. A contact created via API appears in the dashboard. A broadcast sent from the dashboard shows up in the API's campaign stats endpoint.
Read the API docsField Dispatch + Office Coordination
Your dispatch system assigns jobs to field technicians via the API, sending location pins and recording milestones. The office manager monitors site visits from the dashboard in real time, reviewing check-in distances and photo evidence.
API side
- POST /api/v1/messages sends location-based job assignments to technicians
- Webhook captures technician replies with check-in photos
- POST /api/v1/milestones records site_visit_completed events
- Automated alerts fire when a technician misses a check-in window
Dashboard side
- Site Visit tracker shows live who-is-where map
- Office manager reviews check-in distances (self-reported vs expected)
- Photo evidence gallery for completed jobs
- Conversation thread shows full dispatch-to-completion history
Multi-System CRM Sync with Dashboard as Source of Truth
Your external CRM (Salesforce, HubSpot, or custom) syncs contacts bidirectionally via the API. The Klaros dashboard is the canonical view for WhatsApp conversations. Agents see unified history regardless of which system originated the contact.
API side
- POST /api/v1/contacts/bulk imports contacts from external CRM
- Webhooks push new WhatsApp conversations back to external CRM
- PATCH /api/v1/contacts/{id} keeps tags and stages in sync
- GET /api/v1/conversations/{phone}/messages feeds conversation logs to CRM
Dashboard side
- Contact manager shows unified view: CRM-originated and WhatsApp-native contacts
- Agent sees full conversation history regardless of contact source
- Segments built from synced data (CRM tags + WhatsApp engagement)
- Consent status visible and enforced across both systems
Compliance-First Marketing at Scale
Bulk import contacts with consent records via the API. The compliance officer uses the dashboard Compliance tab for audit, CSV export, and right-to-erasure. Every marketing send checks consent programmatically before delivery.
API side
- POST /api/v1/contacts/bulk imports contacts with consent evidence
- POST /api/v1/contacts/{phone}/consent records opt-in with source and evidence
- Campaign API enforces consent check before every send
- Bulk opt-out processing via PATCH endpoints
Dashboard side
- Compliance tab shows consent lookup and record history per contact
- Compliance officer reviews opt-in sources and evidence text
- "Download opted-in (CSV)" exports audit-ready consent evidence
- "Erase permanently" deletes a contact and all messages, logged as proof
Product Catalog Commerce with Conversational Close
Your e-commerce platform syncs the product catalog via the API. Dashboard agents send multi-product messages in conversations, share payment links when the customer is ready, and track conversions in the cost dashboard.
API side
- POST /api/v1/products syncs catalog from e-commerce platform
- Webhook notifies on catalog browsing events
- POST /api/v1/messages sends multi-product list messages
- Product updates (price, stock) sync automatically via API
Dashboard side
- Product library shows the synced catalog with images and prices
- Agent picks products from the library to share in conversation
- Catalog composer builds multi-product messages visually
- Cost dashboard tracks per-conversation revenue attribution
Assistant Books the Meeting, Human Takes Over Cleanly
The built-in assistant handles the moment somebody asks to talk, which is the highest-intent message a business receives, and offers three tappable times without waiting for anyone. The instant a colleague claims that thread, the assistant goes silent and stays silent, so a customer never watches two versions of your business answer the same question.
What the assistant does
- Recognises a request to book, in whatever words and language the customer uses
- Offers up to three times in the customer's own timezone, avoiding what is already booked
- Their tap creates the meeting; nothing is scheduled without a person choosing
- If the diary is full, or it cannot answer, it hands the thread over rather than guessing
What your team sees
- Assign the conversation and the assistant stops replying in it, with no timer to override you
- Answer without assigning and it still stands down for two hours, because the customer is replying to you
- Every decision it made is readable afterwards, including the ones where it chose to stay quiet
- Release the thread and it resumes, so quiet hours are covered without a handover ritual
Start with the dashboard. Extend with the API. Or build API-first and let the team monitor from the dashboard. Same data, same rules, two interfaces.
Start free trialBuild with both surfaces
Same data layer. Same consent rules. Same cost structure. API precision for developers, dashboard visibility for business users.