Faster answers.
Better context.
Blackrock’s customer support agent helps support teams at GCC travel agencies, TMCs and OTAs answer with the booking and platform context already attached. It drafts replies from approved knowledge, in English and Arabic, for a person to review, and hands the issue to a person when it lacks the confidence, permissions or data.
See how it fits: Travel agencies · TMCs · OTAs
An agent with a job, and a person in charge.
Part of Blackrock AI
The customer agent works alongside the CRM, Operations and Intelligence agents.
Humans approve
A person approves anything that moves money or messages a client.
24/7 local support
Local support from our team, around the clock.
From searching to answering.
Speed only helps when the answer is right. Here is what changes when the booking, the knowledge and the hand-over rule sit inside the queue.
TodayA change or cancellation request means opening the booking, the policy and the last conversation in separate systems.
With BlackrockThe booking and platform context arrive with the ticket, where permissions allow.
TodayFinding the right process, product guidance or client terms means knowing where each one is stored.
With BlackrockKnowledge retrieval finds the relevant process, guidance or client-specific information for you.
TodayEvery reply is written from scratch.
With BlackrockA draft from approved knowledge and the booking context is ready for a person to review.
TodaySupport in English and Arabic depends on who is on shift.
With BlackrockDrafts work across English and Arabic, with review and brand tone preserved.
TodayHard cases are escalated by hand, if someone spots them.
With BlackrockThe agent knows when it lacks confidence, permissions or data, and routes the case to a person.
Every ticket arrives prepared.
Booking context, the right knowledge and a draft reply, ready for a person to review.
The booking is already open
Where permissions allow, the agent retrieves the information relevant to the customer’s booking, so nobody has to search across multiple systems.
- Booking summary and status
- The policy and fare rules that apply
- The conversation so far
- Only what the agent’s role is allowed to see
The right process, found for you
Find the relevant internal process, product guidance or client-specific information without knowing where it is stored.
- Internal processes
- Product guidance
- Client-specific information
- Approved knowledge only
Support in English and Arabic, in your tone
Assist support teams working across English and Arabic while preserving review and brand tone.
- Replies drafted in English or Arabic
- Your brand tone preserved
- A person reviews every draft before it is sent
From message to the right next step.
Before it drafts anything, the agent checks what it has: the confidence, the permissions and the data.
It reads
- The message and the conversation so far
- The booking and platform context
- Approved knowledge
Customer support agent
Checks confidence, permissions and data before it acts.
Then it
- Drafts a reply for a person to review
- Hands over to a person
- Asks for approval before money moves or a client is messaged
Ready for faster answers, with the booking attached?
See the customer support agent in a walkthrough with the right team.
Travellers ask on WhatsApp. A person steps in when needed.
Blackrock’s WhatsApp agent answers travellers in the chat they already have open. When the traveller asks for a person, or the request needs a decision, your team takes over with the whole conversation attached.
- Understands the request, including voice notes
- Answers questions about the booking
- Hands over when the traveller asks, or when a decision is needed
- The whole conversation goes with the hand-over
Hands over when a person
should decide.
Support automation should increase trust. Speed is valuable only when the answer is accurate and the user understands what the system can and cannot do.
- Controlled accessBooking information is retrieved only where permissions allow.
- Clear escalationWithout enough confidence, permissions or data, the issue goes to a person.
- Human reviewHumans approve anything that moves money or messages a client.
Start with the workflow, not the model.
We begin with the repetitive decision or information problem in your support queue, then design the agent around the data, tools and controls it needs.
- 01
Pick the problem
Start with the questions your team answers again and again, and the systems they search to answer them.
- 02
Scope the access
Decide what the agent may read, in line with each person’s permissions, and keep sensitive data out of reach.
- 03
Set the hand-over rules
Agree when the agent stops and routes to a person, and which actions a person must approve.
- 04
Go live with review on
Your team reviews drafts and approvals from the start, and agent activity is logged where appropriate.
Delivery is hybrid: you configure the platform yourself while our engineers handle everything specific to your business. Timelines vary with scope, suppliers and custom work.
Questions about AI customer support
Does the agent send replies to customers on its own?
No. It drafts responses using approved knowledge and the available booking context, for a person to review. Humans approve anything that moves money or messages a client.
What can the agent see?
Where permissions allow, it retrieves the information relevant to the customer’s booking. Agent permissions reflect user permissions, access to sensitive data is scoped, and agent activity is logged where appropriate.
Does it work in Arabic?
The agent assists support teams working across English and Arabic while preserving review and brand tone. Our booking products (B2B, B2C and Corporate) support eight languages and full Arabic right-to-left layout.
What happens when the agent is not sure?
It knows when it does not have enough confidence, permissions or data, and routes the issue to a person. On WhatsApp, a person takes over when the traveller asks or when the request needs a decision, with the whole conversation attached.
How is this different from a chatbot?
A separate chatbot is easy to demo and easy to ignore. Blackrock AI works inside the workflows travel teams already use, connected to real business context such as customers, bookings, tasks and conversations.
How do we get started?
We start with the workflow, not the model: the repetitive decision or information problem in your support queue. We then design the agent around the data, tools and controls it needs. Delivery is hybrid, so you configure the platform yourself while our engineers handle everything specific to your business. Timelines vary with scope, suppliers and custom work.