Search & AI Realized
Swirl enterprise search and AI finds and retrieves information from external and internal sources by leveraging traditional search systems. Results are processed with large language models and conversational AI to unlock a wide variety of use cases — from question-answering to insight extraction and workflow automation.
Swirl Deployment Options
Community
$0 Free / Apache 2.0
Community is the base platform for Swirl and available at no cost through an open-source Apache 2.0 license.
Enterprise
Starting at $19/mo/user
Enterprise simplifies deployment as a turnkey per-seat model hosted on Microsoft Azure.
Metapipe
Request pricing
Metapipe connects multiple sources, processes information, and sends to an endpoint.
OPEN-SOURCE ON GITHUB
Swirl Community
Swirl is built on the Python / Django stack, released under the Apache 2.0 license, and available on GitHub. Swirl is available as a Docker image and queries anything with an API then uses spaCy to re-rank the unified results without copying any data!
What Makes Swirl Unique
No Indexing
Swirl adapts and distributes user queries to search engines, databases, and cloud/SaaS services without extracting or indexing anything, saving money and time.
Secure
Swirl enterprise includes OAuth2 support, no need for new permissions. Leverage existing access controls to maintain integrity of source and trusted access to data.
AI Enabled
Swirl uses large language models to help users find and select the most relevant content, and then uses that content to get the richest responses from your AI.
SECURE
Swirl Enterprise
Swirl Enterprise is available in SaaS and Hybird-SaaS offerings on Microsoft Azure. Swirl Enterprise includes connectors to Microsoft 365 Outlook, Teams, Sharepoint, OneDrive and others. Learn more about our growing list of connectors.
Features of Swirl
- No-code query integration
- APIs for Search Providers, Searches and Results
- Synch or asynch metasearching
- Pre-query, query, result and post-result processing pipelines
- Data landed in any Django storage
- Query parsing & adaptation using NLTK
- Matching on word-stems
- Result paging
- Hit highlighting
- Unified results are re-ranked using spaCy LLM
- Result mixers can present results in relevancy, date or round-robin (“stack”) order
- Filtering results to a single source
- Duplicate detection by field or similarity
- Subscribe to update results on schedule
- Expiration service
ENABLE RETRIEVAL AUGMENTED GENERATION (RAG)
Swirl Metapipe
Metapipe builds on the Swirl platform by fetching content from public or private sources, using only existing authentication and permission schemes.
Extract text from numerous formats and use Swirl’s built-in LLM to pre-score and label results, extract entities and sentiment from 12+ languages. Master enriched data in JSON, XML or other for delivery to search engine(s), databases, notifications, indexing, vectorization, summarization, and more.
Metapipe includes processors for chunking, tokenizing content for both Retrieval Augmented Generation (RAG) and fine-tuning of ChatGPT models.
Swirl Pricing
Community
$0 Free / Apache 2.0
Swirl Engine
Swirl user interface
Wiki Documentation
Community Support
Optional support available at $2,500 per month
Enterprise
Starting at $19/mo/user
Enterprise SSO
Access to enterprise connectors
Pre-configured for ChatGPT
Tiered pricing options
Metapipe
Request pricing
Performance monitoring & alerts
Fetch & extract of legacy file formats
Entity & sentiment analysis
Tokenize, score & label w/ Swirl LLM
Retrieval Augmented Generation (RAG) ready with Swirl Enterprise
CONNECTORS
Search and retrieve information from external and internal sources, quickly and securely — without extracting or moving any data.
Swirl is released under the Apache 2.0 license
Copyright © 2023 Swirl, Inc.