Preparing for deployment

The deployment process at Rapidflare is designed to ensure that our agent is fully tailored to the unique needs of each customer. This phase is crucial in setting up the agent to deliver accurate, insightful responses based on the customer's specific products and services. The deployment is entirely cloud-based, allowing for flexibility and quick integration.

Deployment timeline

The exact timeline depends on the complexity and volume of the customer's information sources and any custom features that may need to be configured. Rapidflare's team works closely with customers during this time to ensure that all data is correctly integrated and the agent is aligned with the business requirements.

Teams involved

The deployment process is a collaborative effort between Rapidflare's customer success and engineering teams. The customer success team ensures that the onboarding process goes smoothly, while the engineering team handles any technical configurations or adjustments needed to optimize the agent's performance.

Customization and use case configuration

During deployment, the agent is customized to match the specific use cases of each customer. Common use cases configured by default include:

• Product spec query: Verifying and retrieving technical details for individual products or parts.
• Products lookup query: Searching for products that meet a set of technical specifications.
• Product comparison query: Offering side-by-side comparisons of two or more products, complete with commentary on their similarities and differences.
• Keywords query: Supporting general searches based on keywords or glossary terms, similar to traditional search engines.

In addition, the agent can handle more complex queries such as competitive product comparisons, troubleshooting support, and features related to RFP (Request for Proposal) generation. If user inputs are unclear, the agent is designed to ask follow-up questions or offer suggestions, enhancing the user experience. It can also use past conversation history to suggest autocomplete options or relevant questions.

Testing and validation

Once the agent is configured, the customer success team conducts an initial round of quality assurance (QA). They run various queries through the agent to ensure its performance meets expectations. If any gaps or issues are identified, they are addressed by the engineering team. After this internal testing, a detailed QA report is shared with the customer, showcasing the agent's effectiveness and competitiveness in responding to the customer's unique needs. This thorough process ensures that by the time the agent is deployed, it is fully equipped to handle a range of queries with accuracy and depth, providing immediate value to the customer's business.