22M
~100M
Algolia competitors: EMPATHY vs algolia
Engineer
Sales
Marketing
IP addresses read-only for Platform Engineering operations. No persistence
Offers geo-spatial, geo-location search, as well as IP address capture and persistence
No-Biometric, no-voice, no-image
Offers Voice Search, capturing voice patterns as a unique form of sensitive identifiable information
In-Session IDs only
Uses UserIDs, cross-session identifiers
Not needed, because Empathy is privacy-first
Needed/mandatory
Not needed, because Empathy is privacy-first
Needed/mandatory
Hybrid (keyword & vector) search using ML and NLP. Semantics API
Hybrid (keyword & vector) search using ML and NLP. NeuralSearch API (*only available for clients with the Elevate plan)
Sequential keyword & vector search when searching and indexing. First, keyword search is executed, then vector search if needed.
Parallel keyword & vector search, influencing every search and discovery scenario by dynamically arranging results from most to least relevant.
Open-source foundation semantic models fine-tuned with proprietary datasets for each specific domain.
Language models trained with user-centric and potentially user-held data models, getting the data directly from customers with their consent.
Privacy control measures based on consent-domain integral data as well as semantic models trained to guarantee confidentiality with anonymised in-session behavioural interactions.
Personalisation via different user profiles based on consent, respecting privacy by collecting personal data, not private data.
Play Vector explains through visual insight the what and why behind the AI-generated feed enrichment, semantic models and computed data. Additionally, manual judgments complement offline automated evaluations empowering merchants to orchestrate AI to their ambition.
White-box model of transparency that explains to customers how search relevance is computed, namely, why results are ranked in a particular order.
Self-Hosted Option
SaaS Option
High Availability
RAFT-based multi-node clustering
Multiregion
Only available for self hosted versions
CDN-like Geo-Distributed clusters
Available in Premium tier, called Distributed Search Network
Runtime Dependencies
None. Self-contained binary
Built-in, high performance HTTP server, that can be exposed to the frontend directly
N/A, since it's SaaS only
Index Backward Compatibility
Fully backward compatible
Fully backward compatible
Upgrade Path
Managed SaaS service, doesn't require upgrades
Automated upgrades for self-hosted options based on CI/CD integrations
Managed SaaS service, doesn't require upgrades
Clear separation between search business roles
Equalize/ Static relevancy config (AB)
RT + Curation
NQ + Curation
Promotions (banners)
Redirections
Popular Searches
Top Recommendations
Brand Recommendations (Query Preview)
Next Query Recommendations
Dynamic Re-ranking
Contextualize
Available in Premium Tier
My History
Identifier Results (Sku Search)
Partial Results
Query Field Weights & Boosting (Custom Ranking rules)
Boost/bury products (Custom Ranking rules)
Up to 10 results in Free Tier, Up to 10,000 results in Premium Tier
Explain
Transparency about all aspects that affect the product scoring in order to understand the position of each result in specific queries
Catalog discovery with minimal info about the match (typos, field match, position, etc.) and ranking rules matching
Blacklist of terms
Visual analytics (Insights)
Open statistics/analytics API
Public roadmap - Documentation
Release notes - Documentation
REST API
Typo Tolerance (Spellcheck)
Grouping / Distinct
One distinct field
Dynamic Sorting
Duplicate indices need to be created for each sort order, for Exhaustive Sorting
Virtual replica indices can be used for Relevant Sorting
Faceting & Filtering
Facet Value Searches
Synonyms
Up to 10,000 per index
Dynamic Synonym Suggestions
Available in Premium Tier
Language support
All languages
All languages
Stop words
Geo Search
Search Analytics
Search Success KPIs
Query performance reports
Search Opportunities for improving
Search Performance
Visual analytics covering business cases
Open Statistics API
Client-side and Server-side
Negative Keyword Search (-query)
Phrase Search ("some query")
To Do
User-level Search Personalization
Read-only and in-session / in-the-now for SaaS and read-and-write personalisation for PaaS. More info: Session Contextualization
Premium Tier
A/B Testing Results
Equalize in EPP enables merchandisers to configure and curate the brand search experience by audience segment. Affect product scoring by weighting attribute values and applying them to different segments. Merchandisers have granular control and visibility over how search results appear and the agility to adapt quickly, provide relevant results, and support merchandising goals
Query Suggestions
Wisdom of the Crowd-based
Visual Dashboard
Search & Discovery Playboard
Analytics & Insights for detecting opportunities and understanding shoppers behavior
Tooling for setting up relevance and ranking
Search Experience Customization
Overview of the product grid and scoring details in Explain
Account management
Monitoring of all actions taken in the Playboard
Instance Management Console
Site Crawler
DocSearch crawler for docs sites.
General site crawler is part of Premium Tier
Search UI Component Library
Interface X Components
InstantSearch.js
Privacy Friendliness
Privacy-friendly. Does not collect ANY personal data in either the Self-Hosted nor the SaaS option
Not privacy-friendly. Collects personal data, usage analytics & IP addresses
Number of Documents
Unlimited, only constrained by available RAM
Unknown
Maximum Indices
Unlimited
Unlimited
Maximum Index Size
Unlimited, only constrained by available RAM
Limited to 128 GB
Maximum Words per field
Unlimited
Unlimited
Maximum Record Size
Unlimited
Limited to 10KB
Number of API Keys
Unlimited
Limited to 5000
Channels
GitHub issues
Email
Private Slack (including audion & video calls)
Prioritised Support 24/7
Customer Success specialist
GitHub issues
Email
Public Slack Community
Phone
Paid Prioritized Support
Expert Onboarding & Training
Breadcrumb - Mounted using hierarchical filters
Stats - Number of results
DynamicWidgets - Allows to show/hide widgets depending on the respond. Algolia uses this for Faceting in their examples
Brand Recommendations (Query Preview)
Animations
Currency
Layouts
Modals
Result parts
Scroll
AutoProgressBar
BaseGrid
SlidingPanel
BaseGrid