Ami - the Dialogflow alternative.

Ease of implementation

Dialogflow
Users are required to develop and integrate their own chat widgets, as Dialogflow does not provide one out of the box. Additionally, the platform necessitates a team of programmers to construct and maintain the necessary linguistic structures. Google's support for setting up Dialogflow is limited, often leaving clients to proactively seek out solutions or assistance on their own.

Ami
Ami is fully managed by a team of technical experts based on what the non-technical stakeholders require. Ami's approach makes it quick and easy to deploy an AI without compromising performance and safety.

Safety

Dialogflow
Dialogflow uses Google's in-house generative AI as its core model, and this appears to be the sole option available to users. Because of this, Dialogflow is at risk for hallucination. There is no mention of the anonymisation and the retention of customer data, raising potential concerns about data privacy and security.

Ami
Ami offers prescriptive, generative, and mixed models built and configured locally. Live chat data is not used for training. When using prescriptive and mixed models, Ami completely eliminates hallucinations, making it a risk-free product to use for sensitive clients. When generative technologies are used, Ami's AI framework and its internal QA team aims to eliminate any hallucinations. Ami operates with flexible data retention, meaning that the client is able to choose when the data gets anonymised and for how long it is kept. Generally, all data is kept within the in-house Ami application only and when data gets pushed through a 3rd party API, Ami removes personal information as the conversation is happening. This way no personal data reaches a 3rd party system.


Customisation

Dialogflow
Dialogflow is highly flexible when it comes to integrating with third-party APIs and other systems, allowing businesses to extend its capabilities. However, this flexibility comes with a significant burden on the client, as they must handle the integration work themselves, often with little to no support from Google. This can be a significant barrier for companies without substantial technical resources.

Ami
Ami provides multiple features to give users full control over the conversational experience. The Ami team allows customisation of various components, including the user interface and chatbot behaviour, to the point where the client can choose to control Ami's output word-by-word. Ami’s flexibility extends to integrating CSS for a cohesive brand experience.
 
Extensibility

Dialogflow
The solution is hosted in the client’s help centre. The client themselves can choose the back-end systems they’d like to integrate. However, this is only possible for pre-integrated systems and Zendesk might not offer a wide range of systems to choose from.

Ami
Ami can be integrated to the most common systems in the industry with little to no work. If the required integration is not available out of the box, Ami's integration team will build a custom integration for its clients.

Custom reporting

Dialogflow
Dialogflow includes built-in accuracy reporting as part of its service, which can help businesses assess the performance of the AI in handling customer interactions. However, other conversation data is not provided by default, and Google recommends setting up an integration with another system to capture and analyse this data. Additionally, customer data is only retained for 30 minutes after the last interaction, which could limit the ability to perform more extensive analysis or follow-up actions.

Ami
Ami offers a comprehensive analytics dashboard with over 70 data points and graphs. This extends to conversational analytics but also agent statistics. When requested, Ami's data team can add bespoke insights to any report.