Search operations are content intensive and highly data-driven. For efficient and accurate search results, businesses need to collate content from dependable sources that can cater to their search users’ expectations.
To understand search users’ requirements, businesses must collect, organize and maintain the relevant content and associated parameters from all the diverse sources available.
The Search Assistant lifecycle is structured based on creating and maintaining Search Assistants for your organization.
- Creating Search Assistants and adding relevant content from various sources, defining indices, optimizing results, and tracking the performance, and enterprise Virtual Assistants.
- Maintaining Search Assistants includes
- updating content, and re-defining the search experience for the end-user and deploying onto target channels
- designing and running experiments to test various combinations of configurations to get the maximum value out of your Search Assistants. And if required, based on the outcomes of the experiment fine-tune the indices and repeat the experiments
Create a Search Assistant
Create a customized Search Assistant for your organization to meet your objectives. A Search Assistant is a set of indexing and search configurations applied on a set of content sources along with custom business rules. SearchAssist allows you to build multiple Search Assistants, switch among Search Assistants in the making, launch them and monitor their performance. Refer Managing Search Assistants
Add and Manage Sources
SearchAssist allows you to add and manage content to your Search Assistant from diverse and multiple sources in various formats. Content sources include file uploads, crawling websites with desired frequency, structured data files, FAQs from files or websites and linked virtual assistants. Refer Managing Content.
Manage Indices and Relevance
Once the content has been added, the Search Assistant indexes the data for efficient retrieval. The ingested data from web pages, files, or database records passes through various stages of the indexing process. Indexing management allows you to transform the documents & fields as per the business requirements. Refer Managing Indices.
Manage relevance by fine tuning both the search and results components to refine the search results and present them in meaningful ways. Allow business priorities to be applied and generate custom responses for each query and position certain responses. This helps the end-user to view the most relevant and purposeful results at the same time aligned to the overall business objectives. Refer Managing Relevance and Personalizing Results.
Design Search Experience
Search Experience customization allows you to select the Search Interface between traditional and conversational user interfaces for searching. You can apply customized Result Templates on Results display to enhance the overall search experience. Refer Designing Experience.
Get Analytics and Insights
Analyzing the Search Assistant helps you understand if your search solution is yielding relevant results to your satisfaction. Analyze from the Dashboard and assess the quality of User Engagement from the metrics like Search Insights and Results Insights. These insights further enable you to decide to amend the configuration for delivering better search experience and result outcomes. Refer Analyzing Performance.
Define Experiments
An Experiment helps determine which of different variants or the combination of search and index configurations performs better by presenting each version to live search users at random and analyzing the results. The winner of the experiment is determined by the click-through rate of the variations. Refer Experimenting your Search Variants.
Deployment
Deployment refers to publishing or launching the Search Assistant configuration onto the target channel, a website or an app. This is a two-step process involving adding credentials and activating the channels.