The problem: Great support in complex environments is a multiplayer game

Nearly two years ago Jamie and I started Lorikeet because we observed a massive mismatch between the work of doing support in complex spaces and the FAQ summary-based tools available.

Our conviction was – and remains – that deflection-focused tools that tell customers how to solve problems themselves isn’t enough. We think what’s required is a universal concierge that solves customer problems 24/7 through the channel of their choice by taking actions.

Lorikeet is leading the market in taking actions via first and third party APIs and MCP servers. But across the whole market the gap in action taking is actions that involve other people through phone, SMS, email and Slack. 

Lots of support requires this:

• Financial Services: Checking with the KYC team before resolving onboarding issues

• Healthcare: Calling clinics to reschedule patient appointments or chase consultation notes

• Travel: Calling hotels to arrange late checkouts or special requests

• Creator Economy: Getting manager approval in Slack for discount codes

Until now companies need to give up on resolving these with AI, or attempt the impossible task of fully re-engineering their (and their vendors’) business processes to be API-first.

The solution: A Team of Agents to play the multiplayer game

Since Lorikeet isn’t satisfied with deflection, we stared at this problem and built Team of Agents to solve it. Here’s how it works:

Your main Lorikeet concierge handles the customer conversation. When it needs something from a third party, it spawns a specialized agent to handle that specific task (e.g. calling the doctor, emailing the vendor, texting the delivery driver).

Each spawned agent has clear parameters: what information to gather, what outcome to achieve, and how long to try before escalating. For example, when calling a hotel about a customer request, the agent knows the booking details, the specific request, and that it should wrap up within four minutes (since the customer is waiting).

This works within Lorikeet’s proven framework for safe and controllable AI action-taking. You maintain complete control over how agents operate (natural language instructions or structured workflows), with customizable guardrails for safe, predictable operations.

The result: better customer experiences, and better ROI

The impact of Team of Agents is simple: more delightful customer experiences. This will lead to customer problems getting solved faster and in ways that wow.

As a sweetener it also offers great ROI and greater scalability: the types of issues you can resolve with Team of Agents are meaningfully more expensive than even more complicated procedural tickets as they often involve hand offs to other teams or breaks in workflow to contact third parties.

The possibilities are endless

The part of this launch we’re most excited by isn’t the impact it will have on automating existing manual workloads. It is the way it opens up brand new possibilities. 

There are things businesses simply cannot afford to do today that Team of Agents unlocks. A bank could place a call to a merchant about a dispute before filing the dispute on behalf of their customer. A booking service could call a restaurant for availability as soon as the customer gets in touch with them, instead of handing off.

What will you build?

We love to partner with the most ambitious teams to do cool things for their customers. We’d love to talk to you about what you can build with Team of Agents. Reach out to start building.