Reduce the effort and risk of managing deployments of Fusion in the cloud. Lucidworks has modernized Fusion with a cloud-native microservices architecture orchestrated by Kubernetes. Fusion allows customers to dynamically manage application resources as utilization ebbs and flows, reduce the effort of deploying and upgrading Fusion, and avoid unscheduled downtime and performance degradation.
Fusion includes native support for Python machine learning models. Plug your custom ML models into Fusion and integrate with your existing data science infrastructure. Allow your data scientists and analysts to use Fusion with their existing toolkits to augment user’s human intelligence across the entire experience.
Smart Answers extends the functionality of chatbots and virtual assistants with deep learning. Businesses can better serve their customers and employees by giving them fast answers to their product and support questions.
Predictive Merchandiser gives retailers AI-powered insights and recommendations for optimizing search results and product placement. Merchandisers can apply and manage rules, tune search relevancy, and analyze results through a visual interface without involving IT.
AI-powered search is challenging enough within the English language. The search engine must identify the language, normalize non-English characters, and improve content recall without losing precision. Non-English languages have distinct semantics, grammar and slang, not to mention the unique characters in Asian, European and Arabic base languages. Lucidworks offers an Advanced Linguistics Pack that provides tokenization in more than thirty non-English languages as well as advanced entity extraction for 19 languages.
Cloud-native architectures allow precise management of a platform’s individual software components. Kubernetes orchestrates Fusion’s cloud-native architecture, and sustains the logical separation provided by microservices, the physical separation provided by containers, and the APIs that connect the components. As a result, the Fusion platform can evolve with the changing storage, compute and application ecosystem.
Fusion applies machine learning at the moment of data ingest to add deeper understanding and context to every data set.
Fusion combines the Apache Solr open-source search engine with the distributed power of Apache Spark for artificial intelligence. Highly scalable, the platform indexes and stores data for real-time discovery.
Fusion applies machine learning at query time, to predicts the user intent to return relevant, hyper-personalized results.
Fusion’s application studio allows search engineers to rapidly develop rich applications. It combines an integrated development environment, powerful pre-fabricated components, and APIs for developing powerful search UIs.
Lucidworks offers pricing across three product tiers, either self-hosted by your team or as a Lucidworks managed service in the cloud with list prices calibrated to expected usage levels.
Fusion personalizes queries based on signals like user profile, behavior, natural language processing, or location. Features like autocomplete, query intent classifier, a rules engine, and our proprietary Semantic Knowledge Graph help users express queries for their context. Fusion uses NLP parsing to determine user intent and calculate relevance scores in real-time.
Fusion ships with real-time recommendation algorithms to automatically generate content recommendations based on a user’s past interactions or interactions of similar users.
Continuous feedback loops help domain experts optimize machine learning results combined with manual rules curation.
Fusion’s application studio lets you quickly create production-ready applications. Combine powerful pre-fabricated components and APIs for developing powerful search UIs.
Analytic insights lets you graphically analyze user behavior, run experiments (such as A/B testing), and inspect individual customer journeys.
Fusion employs NLP to detect phrases, topics and parts of speech for automatic classification and clustering of content in ways that make it easily accessible for search, browsing and predictive suggestions.
Fusion clusters users by behavior and elevates documents which perform best for each cluster’s favorite queries. Models then boost documents with features most relevant to a given cluster.
Head/Tail Analysis and automatic synonym detection analyzes the head (most common) and tail (infrequent) queries in the system, then auto-generates synonyms.
Machine learning in Fusion uses training data on past searches and outcomes to predict the intent of each new query. Deep learning can use that intelligence for query parsing, query pipeline routing, autocomplete and type-ahead.
The Learning to Rank algorithm extracts tags such as product names, titles, and document categories to determine relevance scores in real-time.
Fusion uses unsupervised machine learning to cluster similar content for pattern recognition. Classification, a supervised form of ML, teaches Fusion to classify new data to predetermined categories.