Building an AI Agent inside Amazon Connect

Learn how HOP automated customer support, unified communications, and gained greater control over its network with Amazon Connect and AI.

About HOP

HOP logo

About HOP

HOP is a logistics company that operates a nationwide network of hundreds of pickup points, connecting businesses and customers through a distributed last-mile delivery infrastructure.

Its operation depends on efficient, scalable communication with locations across the network to manage package status, operational updates, onboarding, and day-to-day support.

The Challenge

For companies operating through a broad network of intermediaries or providers—distributors, points of sale, pickup locations, franchises—communication is one of the biggest operational bottlenecks. Stock notifications, status updates, document delivery, and responses to inquiries all generate a volume of interactions that is difficult to scale without sacrificing quality or overwhelming the support team.
Many of these companies have already addressed part of the problem by adopting Amazon Connect as their contact center platform. Their flows are configured, their agents work from there, and their communication channels are unified. However, they often face one specific limitation: Connect centralizes inbound support, but it does not provide a simple, direct way to handle outbound communication with providers, nor does it reduce the burden on human teams when dealing with repetitive, operational inquiries.
As a result, teams end up working across parallel tools: a spreadsheet for bulk sends, another system for WhatsApp campaigns, and the contact center for inbound interactions. This creates fragmentation, rework, and dependence on technical profiles for tasks that operators should be able to handle themselves.

DinoCloud’s proposed solution

When HOP came to DinoCloud with this challenge, we designed a solution that did not replace Amazon Connect, but enhanced it from within.
The solution was built around two pillars:

  1. Integration of a conversational AI agent (Amazon Bedrock). A Bedrock agent was integrated directly into Connect flows, enabling it to autonomously answer operational inquiries from HOP locations, including package status, onboarding requests, and updates. The agent understands the business because it was trained on HOP’s own rules and operational logic. If an inquiry exceeds its ability to resolve it, the conversation is automatically escalated to a human agent, always within the same Connect environment.
  2. Custom outbound communication modules integrated into Connect. Two capabilities were developed to meet HOP’s needs because Connect’s native bulk WhatsApp messaging functionality did not fit the client’s existing workflows:
    Bulk WhatsApp campaign sending: the operator uploads a CSV file with the phone numbers of HOP locations and selects a message template. The campaign is sent directly from Connect, without external tools or technical intervention.
    Opening a 1-to-1 support channel: when an operator needs to start a conversation with a specific HOP location, they can do so directly from their workspace. An initial template message is sent—as required by Meta to open the conversation—the human support window is enabled, and once the operator closes it, the channel automatically returns to bot handling.

At DinoCloud, we specialize in building AI solutions that deliver real business outcomes. From automating customer interactions to streamlining operations and enabling teams to work more efficiently, we design solutions that turn AI into measurable impact. If you’re looking to apply AI to a real business challenge, feel free to contact us here.

How we built it

Development of the custom modules took approximately two weeks. The technical key was building small applications embedded directly within the Connect interface, so that from the operator’s perspective they look and work as if they had always been part of the platform. There are no external windows, no system switching, and no need to access the AWS console.
This is possible because Amazon Connect allows its interface to be extended through applications integrated into the agent workspace. DinoCloud used this capability to add real business functionality—not just configuration—directly into the environment where the team already works.

Results

Even without waiting for formal metrics, the architecture of this solution has concrete and predictable operational impacts:

Why this matters if you already use Amazon Connect

If Amazon Connect is already running in your contact center, DinoCloud can extend it with custom modules tailored to your operation: outbound communication, bulk uploads, purpose-built dashboards, and integrations with your business logic. All without leaving the platform, adding tools to the stack, or launching months-long projects.

It’s a quick win with a real impact on your team’s day-to-day operations.

We are #YourIdealCloudPartner

Focus on your core business while DinoCloud provides the
technology outcomes you need by leveraging its expertise in
the cloud.

Get in touch

(*) Required fields