How Stellar works

Stellar's platform supports all phases of the agent development lifecycle. If needed, our team can support you across all phases from initial agent design to continued support and maintenance

1. Create your AI agent

First determine your agent's persona.You can select from a range of voices, tones and personalities to match your brand preference.

Next, instruct your agent how to resolve different types of customer requests by giving it dedicated instructions per conversation topic. Stellar's Dialogue Builder makes it easy and intuitive to provide specific instructions for each individual type of request.

Stellar can initially take care of full configuration for your agent, based on your exact business needs. Your employees can receive training and continued support in managing Stellar agents, so that they can learn to improve and manage your AI agents independently.

2. Enhance your agent with actions & knowledge

Connect the agent with your external systems via secure API integrations, enabling it to safely gather specific information and take pre-approved specific actions for your customers. Stellar can integrate with any system that exposes an API.

Next, enhance your agent with the knowledge needed to properly assist customers, by feeding it your processes and other internal documentation. If needed, our team can help optimise your existing documentation for processing by the agent, to enhance accuracy and reliability.

3. Set the boundary conditions & limits

Use our guardrail system to mark certain topics and situations as off-limits for your AI agents, and to set clear rules that are enforced at every turn of the conversation for added safety.

Define when and how to transfer calls to your human support agents.
‍
Stellar has developed a set of guardrails for common scenarios to use as a safe starting point, and on top of that we help you define your own additional business-specific guardrails.

4. Test your agent

To make sure your agent is and remains robust, we allow you to test various parts of your agent throughout all stages of the agent development lifecycle.

During agent configuration, Stellar's Playground allows instant test calls with your agent to easily iterate.

Use our automated test suites to quickly test for regressions across a wide array of scenarios. This ensures that your recent changes don't break anything.

Post-production monitoring and optional automated smoke tests allow staying on top of agent performance during conversations with your actual customers.

5. Deploy to production

Expose your agent to the outside world, either by connecting it with your existing phone support numbers or
by embedding the agent on your website / web app.

By default your agents run on Stellar's cloud infrastructure. Alternative deployment setups can be discussed if required by your use case or policies.

Stellar offers deep integration with callcenter software providers, allowing our agents to act as any other support agent in your call queues. Integrate with Stellar via API (via bi-directional audio streaming) or via SIP trunking. Decide on a per-topic basis whether to route to Stellar's agents or your human representatives (Stellar can always transfer customers back to your human representatives where needed)

6. Monitor & continuously improve

Effectively monitor agent performance, resolution rates, hand-offs, and trending topics using our pre-configured monitoring dashboards. Easily drill down on potential opportunities through call recordings, transcripts, summaries and built-in analytics.

Define your own compliance checks to maintain compliant with regulation and policies. Receive alerts and reports based on your preferences.

If preferred, Stellar can sync monitoring data and reporting to your own data warehouse and/or BI solution.

Easily track performance and opportunities

Easily stay on track of how your Stellar AI agents are performing and where opportunities for improvement lie. Our platform gives you all the tools and insights you need.

Experience our white glove service

Robustly implementing AI is challenging. You know your business better than anyone, but you are likely not an expert on AI voice agents. We are.

Our team can help configure your initial AI voice agent, and guide you through every step along the way to production while closely collaborating with you and your team. Leverage our hard-earned learnings when it comes to successfully launching voice AI in enterprise and highly regulated settings.

We focus on long-term partnerships, and can act as your partner from strategic prioritisation through to project planning, agent design, and integration with your existing IT landscape. Start small, and expand iteratively.

Effectively launch your AI voice agents to production